Python tf_util.conv2d() Examples

def build(self,
            inputs,
            num_outputs,
            scope=None,
            activation_fn=tf.nn.relu,
            is_training=None):
    ''' Build Multi-layer preceptrons '''
    outputs = tf_util.conv2d(inputs,
                             num_outputs,
                             self.kernel_size,
                             padding=self.padding,
                             stride=self.stride,
                             bn=self.bn,
                             is_training=is_training,
                             weight_decay=self.weight_decay,
                             activation_fn = activation_fn,
                             scope=scope,
                             bn_decay=self.bn_decay,
                             is_dist=self.is_dist)

    return outputs 

def residual_block(X, out_dim, is_training, bn_decay, scope):
    n_dim = X.get_shape()[2].value
    X_expanded = tf.expand_dims(X, -2)

    net = tf_util.conv2d(X_expanded, out_dim, [1,1], padding='VALID',
            stride=[1,1], bn=True, is_training=is_training,
            bn_decay=bn_decay, activation_fn=tf.nn.relu, scope=scope+'_conv1')

    net = tf_util.conv2d(net, out_dim, [1,1], padding='VALID',
            stride=[1,1], bn=True, is_training=is_training,
            bn_decay=bn_decay, activation_fn=None, scope=scope+'_conv2')

    if n_dim != out_dim:
        X_expanded = tf_util.conv2d(X_expanded, out_dim, [1,1], padding='VALID',
                stride=[1,1], activation_fn=None, scope=scope+'_conv3')

    net = tf.nn.relu(net + X_expanded)
    net = tf.squeeze(net)

    return net 

def single_encoding_net(pc, mlp_list, mlp_list2, scope, is_training, bn_decay):
    ''' The encoding network for instance
    Input:
        pc: [B, N, 3]
    Return:
        net: [B, nfea]
    '''
    with tf.variable_scope(scope) as myscope:
        net = tf.expand_dims(pc, 2)
        for i,num_out_channel in enumerate(mlp_list):
            net = tf_util.conv2d(net, num_out_channel, [1,1],
                                 padding='VALID', stride=[1,1],
                                 bn=True, is_training=is_training,
                                 scope='conv%d'%i, bn_decay=bn_decay)
        net = tf.reduce_max(net, axis=[1])
        net = tf.squeeze(net, 1)
        for i,num_out_channel in enumerate(mlp_list2):
            net = tf_util.fully_connected(net, num_out_channel, bn=True, is_training=is_training,
                                          scope='fc%d'%i, bn_decay=bn_decay)
        return net 

def weight_net(xyz, hidden_units, scope, is_training, bn_decay=None, weight_decay=None,
               activation_fn=tf.nn.relu, is_dist=False):
    with tf.variable_scope(scope) as sc:
        net = xyz
        for i, num_hidden_units in enumerate(hidden_units):
            if i != len(hidden_units) - 1:
                net = tf_util.conv2d(net, num_hidden_units, [1, 1],
                                     padding='VALID', stride=[1, 1],
                                     bn=True, is_training=is_training, activation_fn=activation_fn,
                                     scope='wconv{}'.format(i), bn_decay=bn_decay,
                                     weight_decay=weight_decay, is_dist=is_dist)
            else:
                net = tf_util.conv2d(net, num_hidden_units, [1, 1],
                                     padding='VALID', stride=[1, 1],
                                     bn=False, is_training=is_training, activation_fn=None,
                                     scope='wconv{}'.format(i), bn_decay=bn_decay,
                                     weight_decay=weight_decay, is_dist=is_dist)
            # net = tf_util.dropout(net, keep_prob=0.5, is_training=is_training, scope='wconv_dp{}'.format(i))
    return net 

def nonlinear_transform(data_in, mlp, scope, is_training, bn_decay=None, weight_decay=None,
                        activation_fn=tf.nn.relu, is_dist=False):
    with tf.variable_scope(scope) as sc:

        net = data_in
        l = len(mlp)
        if l > 1:
            for i, out_ch in enumerate(mlp[0:(l - 1)]):
                net = tf_util.conv2d(net, out_ch, [1, 1],
                                     padding='VALID', stride=[1, 1],
                                     bn=True, is_training=is_training, activation_fn=tf.nn.relu,
                                     scope='nonlinear{}'.format(i), bn_decay=bn_decay,
                                     weight_decay=weight_decay, is_dist=is_dist)

                # net = tf_util.dropout(net, keep_prob=0.5, is_training=is_training, scope='dp_nonlinear{}'.format(i))
        net = tf_util.conv2d(net, mlp[-1], [1, 1],
                             padding='VALID', stride=[1, 1],
                             bn=False, is_training=is_training,
                             scope='nonlinear%d' % (l - 1), bn_decay=bn_decay,
                             activation_fn=tf.nn.sigmoid, weight_decay=weight_decay, is_dist=is_dist)

    return net 

def _get_dgcnn(pcs, layer_sizes, scope_name, is_training, bn_decay):
    assert len(layer_sizes) > 0
    num_point = pcs.shape[1]
    k = 20
    with tf.variable_scope(scope_name):
        adj_matrix = tf_util_dgcnn.pairwise_distance(pcs)
        nn_idx = tf_util_dgcnn.knn(adj_matrix, k=k)
        edge_feature = tf_util_dgcnn.get_edge_feature(pcs, nn_idx=nn_idx, k=k)

        net = tf_util_dgcnn.conv2d(edge_feature, layer_sizes[0], [1, 1], padding='VALID', stride=[1, 1], bn=True, is_training=is_training, scope='conv1', bn_decay=bn_decay)
        for idx, layer_size in enumerate(layer_sizes[1:-1]):
            net = tf_util_dgcnn.conv2d(net, layer_size, [1, 1], padding='VALID', stride=[1, 1], bn=True, is_training=is_training, scope=f'conv{idx+2}', bn_decay=bn_decay)
        net = tf.reduce_max(net, axis=-2, keepdims=True)
        net = tf_util_dgcnn.conv2d(net, layer_sizes[-1], [1, 1], padding='VALID', stride=[1, 1], bn=True, is_training=is_training, scope=f'conv{len(layer_sizes)}', bn_decay=bn_decay)

        net = tf_util_dgcnn.max_pool2d(net, [num_point, 1], padding='VALID', scope='maxpool')
        return net 

def get_center_regression_net(object_point_cloud, one_hot_vec,
                              is_training, bn_decay, end_points):
    ''' Regression network for center delta. a.k.a. T-Net.
    Input:
        object_point_cloud: TF tensor in shape (B,M,C)
            point clouds in 3D mask coordinate
        one_hot_vec: TF tensor in shape (B,3)
            length-3 vectors indicating predicted object type
    Output:
        predicted_center: TF tensor in shape (B,3)
    ''' 
    num_point = object_point_cloud.get_shape()[1].value
    net = tf.expand_dims(object_point_cloud, 2)
    net = tf_util.conv2d(net, 128, [1,1],
                         padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training,
                         scope='conv-reg1-stage1', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 128, [1,1],
                         padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training,
                         scope='conv-reg2-stage1', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 256, [1,1],
                         padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training,
                         scope='conv-reg3-stage1', bn_decay=bn_decay)
    net = tf_util.max_pool2d(net, [num_point,1],
        padding='VALID', scope='maxpool-stage1')
    net = tf.squeeze(net, axis=[1,2])
    net = tf.concat([net, one_hot_vec], axis=1)
    net = tf_util.fully_connected(net, 256, scope='fc1-stage1', bn=True,
        is_training=is_training, bn_decay=bn_decay)
    net = tf_util.fully_connected(net, 128, scope='fc2-stage1', bn=True,
        is_training=is_training, bn_decay=bn_decay)
    predicted_center = tf_util.fully_connected(net, 3, activation_fn=None,
        scope='fc3-stage1')
    return predicted_center, end_points 

def get_transform(point_cloud, is_training, bn_decay=None, K = 3):
    """ Transform Net, input is BxNx3 gray image
        Return:
            Transformation matrix of size 3xK """
    batch_size = point_cloud.get_shape()[0].value
    num_point = point_cloud.get_shape()[1].value

    input_image = tf.expand_dims(point_cloud, -1)
    net = tf_util.conv2d(input_image, 64, [1,3], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='tconv1', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 128, [1,1], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='tconv3', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 1024, [1,1], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='tconv4', bn_decay=bn_decay)
    net = tf_util.max_pool2d(net, [num_point,1], padding='VALID', scope='tmaxpool')

    net = tf.reshape(net, [batch_size, -1])
    net = tf_util.fully_connected(net, 128, bn=True, is_training=is_training, scope='tfc1', bn_decay=bn_decay)
    net = tf_util.fully_connected(net, 128, bn=True, is_training=is_training, scope='tfc2', bn_decay=bn_decay)

    with tf.variable_scope('transform_XYZ') as sc:
        assert(K==3)
        weights = tf.get_variable('weights', [128, 3*K], initializer=tf.constant_initializer(0.0), dtype=tf.float32)
        biases = tf.get_variable('biases', [3*K], initializer=tf.constant_initializer(0.0), dtype=tf.float32) + tf.constant([1,0,0,0,1,0,0,0,1], dtype=tf.float32)
        transform = tf.matmul(net, weights)
        transform = tf.nn.bias_add(transform, biases)

    #transform = tf_util.fully_connected(net, 3*K, activation_fn=None, scope='tfc3')
    transform = tf.reshape(transform, [batch_size, 3, K])
    return transform 

def pointnet_sa_module_msg(xyz, points, npoint, radius_list, nsample_list, mlp_list, is_training, bn_decay, scope, bn=True, use_xyz=True, use_nchw=False):
    ''' PointNet Set Abstraction (SA) module with Multi-Scale Grouping (MSG)
        Input:
            xyz: (batch_size, ndataset, 3) TF tensor
            points: (batch_size, ndataset, channel) TF tensor
            npoint: int32 -- #points sampled in farthest point sampling
            radius: list of float32 -- search radius in local region
            nsample: list of int32 -- how many points in each local region
            mlp: list of list of int32 -- output size for MLP on each point
            use_xyz: bool, if True concat XYZ with local point features, otherwise just use point features
            use_nchw: bool, if True, use NCHW data format for conv2d, which is usually faster than NHWC format
        Return:
            new_xyz: (batch_size, npoint, 3) TF tensor
            new_points: (batch_size, npoint, \sum_k{mlp[k][-1]}) TF tensor
    '''
    data_format = 'NCHW' if use_nchw else 'NHWC'
    with tf.variable_scope(scope) as sc:
        new_xyz = gather_point(xyz, farthest_point_sample(npoint, xyz))
        new_points_list = []
        for i in range(len(radius_list)):
            radius = radius_list[i]
            nsample = nsample_list[i]
            idx, pts_cnt = query_ball_point(radius, nsample, xyz, new_xyz)
            grouped_xyz = group_point(xyz, idx)
            grouped_xyz -= tf.tile(tf.expand_dims(new_xyz, 2), [1,1,nsample,1])
            if points is not None:
                grouped_points = group_point(points, idx)
                if use_xyz:
                    grouped_points = tf.concat([grouped_points, grouped_xyz], axis=-1)
            else:
                grouped_points = grouped_xyz
            if use_nchw: grouped_points = tf.transpose(grouped_points, [0,3,1,2])
            for j,num_out_channel in enumerate(mlp_list[i]):
                grouped_points = tf_util.conv2d(grouped_points, num_out_channel, [1,1],
                                                padding='VALID', stride=[1,1], bn=bn, is_training=is_training,
                                                scope='conv%d_%d'%(i,j), bn_decay=bn_decay)
            if use_nchw: grouped_points = tf.transpose(grouped_points, [0,2,3,1])
            new_points = tf.reduce_max(grouped_points, axis=[2])
            new_points_list.append(new_points)
        new_points_concat = tf.concat(new_points_list, axis=-1)
        return new_xyz, new_points_concat 

def pointnet_fp_module(xyz1, xyz2, points1, points2, mlp, is_training, bn_decay, scope, bn=True):
    ''' PointNet Feature Propogation (FP) Module
        Input:                                                                                                      
            xyz1: (batch_size, ndataset1, 3) TF tensor                                                              
            xyz2: (batch_size, ndataset2, 3) TF tensor, sparser than xyz1                                           
            points1: (batch_size, ndataset1, nchannel1) TF tensor                                                   
            points2: (batch_size, ndataset2, nchannel2) TF tensor
            mlp: list of int32 -- output size for MLP on each point                                                 
        Return:
            new_points: (batch_size, ndataset1, mlp[-1]) TF tensor
    '''
    with tf.variable_scope(scope) as sc:
        dist, idx = three_nn(xyz1, xyz2)
        dist = tf.maximum(dist, 1e-10)
        norm = tf.reduce_sum((1.0/dist),axis=2,keep_dims=True)
        norm = tf.tile(norm,[1,1,3])
        weight = (1.0/dist) / norm
        interpolated_points = three_interpolate(points2, idx, weight)

        if points1 is not None:
            new_points1 = tf.concat(axis=2, values=[interpolated_points, points1]) # B,ndataset1,nchannel1+nchannel2
        else:
            new_points1 = interpolated_points
        new_points1 = tf.expand_dims(new_points1, 2)
        for i, num_out_channel in enumerate(mlp):
            new_points1 = tf_util.conv2d(new_points1, num_out_channel, [1,1],
                                         padding='VALID', stride=[1,1],
                                         bn=bn, is_training=is_training,
                                         scope='conv_%d'%(i), bn_decay=bn_decay)
        new_points1 = tf.squeeze(new_points1, [2]) # B,ndataset1,mlp[-1]
        return new_points1 

def pointnet_sa_module_msg(xyz, points, npoint, radius_list, nsample_list, mlp_list, is_training, bn_decay, scope, bn=True, use_xyz=True, use_nchw=False):
    ''' PointNet Set Abstraction (SA) module with Multi-Scale Grouping (MSG)
        Input:
            xyz: (batch_size, ndataset, 3) TF tensor
            points: (batch_size, ndataset, channel) TF tensor
            npoint: int32 -- #points sampled in farthest point sampling
            radius: list of float32 -- search radius in local region
            nsample: list of int32 -- how many points in each local region
            mlp: list of list of int32 -- output size for MLP on each point
            use_xyz: bool, if True concat XYZ with local point features, otherwise just use point features
            use_nchw: bool, if True, use NCHW data format for conv2d, which is usually faster than NHWC format
        Return:
            new_xyz: (batch_size, npoint, 3) TF tensor
            new_points: (batch_size, npoint, \sum_k{mlp[k][-1]}) TF tensor
    '''
    data_format = 'NCHW' if use_nchw else 'NHWC'
    with tf.variable_scope(scope) as sc:
        new_xyz = gather_point(xyz, farthest_point_sample(npoint, xyz))
        new_points_list = []
        for i in range(len(radius_list)):
            radius = radius_list[i]
            nsample = nsample_list[i]
            idx, pts_cnt = query_ball_point(radius, nsample, xyz, new_xyz)
            grouped_xyz = group_point(xyz, idx)
            grouped_xyz -= tf.tile(tf.expand_dims(new_xyz, 2), [1,1,nsample,1])
            if points is not None:
                grouped_points = group_point(points, idx)
                if use_xyz:
                    grouped_points = tf.concat([grouped_points, grouped_xyz], axis=-1)
            else:
                grouped_points = grouped_xyz
            if use_nchw: grouped_points = tf.transpose(grouped_points, [0,3,1,2])
            for j,num_out_channel in enumerate(mlp_list[i]):
                grouped_points = tf_util.conv2d(grouped_points, num_out_channel, [1,1],
                                                padding='VALID', stride=[1,1], bn=bn, is_training=is_training,
                                                scope='conv%d_%d'%(i,j), bn_decay=bn_decay)
            if use_nchw: grouped_points = tf.transpose(grouped_points, [0,2,3,1])
            new_points = tf.reduce_max(grouped_points, axis=[2])
            new_points_list.append(new_points)
        new_points_concat = tf.concat(new_points_list, axis=-1)
        return new_xyz, new_points_concat 

def pointnet_fp_module(xyz1, xyz2, points1, points2, mlp, is_training, bn_decay, scope, bn=True, is_dist=False):
    ''' PointNet Feature Propogation (FP) Module
        Input:                                                                                                      
            xyz1: (batch_size, ndataset1, 3) TF tensor                                                              
            xyz2: (batch_size, ndataset2, 3) TF tensor, sparser than xyz1                                           
            points1: (batch_size, ndataset1, nchannel1) TF tensor                                                   
            points2: (batch_size, ndataset2, nchannel2) TF tensor
            mlp: list of int32 -- output size for MLP on each point                                                 
        Return:
            new_points: (batch_size, ndataset1, mlp[-1]) TF tensor
    '''
    with tf.variable_scope(scope) as sc:
        dist, idx = three_nn(xyz1, xyz2)
        dist = tf.maximum(dist, 1e-10)
        norm = tf.reduce_sum((1.0 / dist), axis=2, keep_dims=True)
        norm = tf.tile(norm, [1, 1, 3])
        weight = (1.0 / dist) / norm
        interpolated_points = three_interpolate(points2, idx, weight)

        if points1 is not None:
            new_points1 = tf.concat(axis=2, values=[interpolated_points, points1])  # B,ndataset1,nchannel1+nchannel2
        else:
            new_points1 = interpolated_points
        new_points1 = tf.expand_dims(new_points1, 2)
        for i, num_out_channel in enumerate(mlp):
            new_points1 = tf_util.conv2d(new_points1, num_out_channel, [1, 1],
                                         padding='VALID', stride=[1, 1],
                                         bn=bn, is_training=is_training,
                                         scope='conv_%d' % (i), bn_decay=bn_decay,
                                         is_dist=is_dist)
        new_points1 = tf.squeeze(new_points1, [2])  # B,ndataset1,mlp[-1]
        return new_points1 

def get_transform(point_cloud, is_training, bn_decay=None, K = 3):
    """ Transform Net, input is BxNx3 gray image
        Return:
            Transformation matrix of size 3xK """
    batch_size = point_cloud.get_shape()[0].value
    num_point = point_cloud.get_shape()[1].value

    input_image = tf.expand_dims(point_cloud, -1)
    net = tf_util.conv2d(input_image, 64, [1,3], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='tconv1', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 128, [1,1], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='tconv3', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 1024, [1,1], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='tconv4', bn_decay=bn_decay)
    net = tf_util.max_pool2d(net, [num_point,1], padding='VALID', scope='tmaxpool')

    net = tf.reshape(net, [batch_size, -1])
    net = tf_util.fully_connected(net, 128, bn=True, is_training=is_training, scope='tfc1', bn_decay=bn_decay)
    net = tf_util.fully_connected(net, 128, bn=True, is_training=is_training, scope='tfc2', bn_decay=bn_decay)

    with tf.variable_scope('transform_XYZ') as sc:
        assert(K==3)
        weights = tf.get_variable('weights', [128, 3*K], initializer=tf.constant_initializer(0.0), dtype=tf.float32)
        biases = tf.get_variable('biases', [3*K], initializer=tf.constant_initializer(0.0), dtype=tf.float32) + tf.constant([1,0,0,0,1,0,0,0,1], dtype=tf.float32)
        transform = tf.matmul(net, weights)
        transform = tf.nn.bias_add(transform, biases)

    #transform = tf_util.fully_connected(net, 3*K, activation_fn=None, scope='tfc3')
    transform = tf.reshape(transform, [batch_size, 3, K])
    return transform 

def get_transform_K(inputs, is_training, bn_decay=None, K = 3):
    """ Transform Net, input is BxNx1xK gray image
        Return:
            Transformation matrix of size KxK """
    batch_size = inputs.get_shape()[0].value
    num_point = inputs.get_shape()[1].value

    net = tf_util.conv2d(inputs, 256, [1,1], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='tconv1', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 1024, [1,1], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='tconv2', bn_decay=bn_decay)
    net = tf_util.max_pool2d(net, [num_point,1], padding='VALID', scope='tmaxpool')

    net = tf.reshape(net, [batch_size, -1])
    net = tf_util.fully_connected(net, 512, bn=True, is_training=is_training, scope='tfc1', bn_decay=bn_decay)
    net = tf_util.fully_connected(net, 256, bn=True, is_training=is_training, scope='tfc2', bn_decay=bn_decay)

    with tf.variable_scope('transform_feat') as sc:
        weights = tf.get_variable('weights', [256, K*K], initializer=tf.constant_initializer(0.0), dtype=tf.float32)
        biases = tf.get_variable('biases', [K*K], initializer=tf.constant_initializer(0.0), dtype=tf.float32) + tf.constant(np.eye(K).flatten(), dtype=tf.float32)
        transform = tf.matmul(net, weights)
        transform = tf.nn.bias_add(transform, biases)

    #transform = tf_util.fully_connected(net, 3*K, activation_fn=None, scope='tfc3')
    transform = tf.reshape(transform, [batch_size, K, K])
    return transform 

def pointnet_sa_module_msg(xyz, points, npoint, radius_list, nsample_list, mlp_list, is_training, bn_decay, scope, bn=True, use_xyz=True, use_nchw=False):
    ''' PointNet Set Abstraction (SA) module with Multi-Scale Grouping (MSG)
        Input:
            xyz: (batch_size, ndataset, 3) TF tensor
            points: (batch_size, ndataset, channel) TF tensor
            npoint: int32 -- #points sampled in farthest point sampling
            radius: list of float32 -- search radius in local region
            nsample: list of int32 -- how many points in each local region
            mlp: list of list of int32 -- output size for MLP on each point
            use_xyz: bool, if True concat XYZ with local point features, otherwise just use point features
            use_nchw: bool, if True, use NCHW data format for conv2d, which is usually faster than NHWC format
        Return:
            new_xyz: (batch_size, npoint, 3) TF tensor
            new_points: (batch_size, npoint, \sum_k{mlp[k][-1]}) TF tensor
    '''
    data_format = 'NCHW' if use_nchw else 'NHWC'
    with tf.variable_scope(scope) as sc:
        new_xyz = gather_point(xyz, farthest_point_sample(npoint, xyz))
        new_points_list = []
        for i in range(len(radius_list)):
            radius = radius_list[i]
            nsample = nsample_list[i]
            idx, pts_cnt = query_ball_point(radius, nsample, xyz, new_xyz)
            grouped_xyz = group_point(xyz, idx)
            grouped_xyz -= tf.tile(tf.expand_dims(new_xyz, 2), [1,1,nsample,1])
            if points is not None:
                grouped_points = group_point(points, idx)
                if use_xyz:
                    grouped_points = tf.concat([grouped_points, grouped_xyz], axis=-1)
            else:
                grouped_points = grouped_xyz
            if use_nchw: grouped_points = tf.transpose(grouped_points, [0,3,1,2])
            for j,num_out_channel in enumerate(mlp_list[i]):
                grouped_points = tf_util.conv2d(grouped_points, num_out_channel, [1,1],
                                                padding='VALID', stride=[1,1], bn=bn, is_training=is_training,
                                                scope='conv%d_%d'%(i,j), bn_decay=bn_decay)
            if use_nchw: grouped_points = tf.transpose(grouped_points, [0,2,3,1])
            new_points = tf.reduce_max(grouped_points, axis=[2])
            new_points_list.append(new_points)
        new_points_concat = tf.concat(new_points_list, axis=-1)
        return new_xyz, new_points_concat 

def pointnet_sa_module_msg(xyz, points, npoint, radius_list, nsample_list, mlp_list, is_training, bn_decay, scope, bn=True, use_xyz=True, use_nchw=False):
    ''' PointNet Set Abstraction (SA) module with Multi-Scale Grouping (MSG)
        Input:
            xyz: (batch_size, ndataset, 3) TF tensor
            points: (batch_size, ndataset, channel) TF tensor
            npoint: int32 -- #points sampled in farthest point sampling
            radius: list of float32 -- search radius in local region
            nsample: list of int32 -- how many points in each local region
            mlp: list of list of int32 -- output size for MLP on each point
            use_xyz: bool, if True concat XYZ with local point features, otherwise just use point features
            use_nchw: bool, if True, use NCHW data format for conv2d, which is usually faster than NHWC format
        Return:
            new_xyz: (batch_size, npoint, 3) TF tensor
            new_points: (batch_size, npoint, \sum_k{mlp[k][-1]}) TF tensor
    '''
    data_format = 'NCHW' if use_nchw else 'NHWC'
    with tf.variable_scope(scope) as sc:
        new_xyz = gather_point(xyz, farthest_point_sample(npoint, xyz))
        new_points_list = []
        for i in range(len(radius_list)):
            radius = radius_list[i]
            nsample = nsample_list[i]
            idx, pts_cnt = query_ball_point(radius, nsample, xyz, new_xyz)
            grouped_xyz = group_point(xyz, idx)
            grouped_xyz -= tf.tile(tf.expand_dims(new_xyz, 2), [1,1,nsample,1])
            if points is not None:
                grouped_points = group_point(points, idx)
                if use_xyz:
                    grouped_points = tf.concat([grouped_points, grouped_xyz], axis=-1)
            else:
                grouped_points = grouped_xyz
            if use_nchw: grouped_points = tf.transpose(grouped_points, [0,3,1,2])
            for j,num_out_channel in enumerate(mlp_list[i]):
                grouped_points = tf_util.conv2d(grouped_points, num_out_channel, [1,1],
                                                padding='VALID', stride=[1,1], bn=bn, is_training=is_training,
                                                scope='conv%d_%d'%(i,j), bn_decay=bn_decay)
            if use_nchw: grouped_points = tf.transpose(grouped_points, [0,2,3,1])
            new_points = tf.reduce_max(grouped_points, axis=[2])
            new_points_list.append(new_points)
        new_points_concat = tf.concat(new_points_list, axis=-1)
        return new_xyz, new_points_concat 

def pointnet_fp_module(xyz1, xyz2, points1, points2, mlp, is_training, bn_decay, scope, bn=True, last_mlp_activation=True):
    ''' PointNet Feature Propogation (FP) Module
        Input:
            xyz1: (batch_size, ndataset1, 3) TF tensor
            xyz2: (batch_size, ndataset2, 3) TF tensor, sparser than xyz1
            points1: (batch_size, ndataset1, nchannel1) TF tensor
            points2: (batch_size, ndataset2, nchannel2) TF tensor
            mlp: list of int32 -- output size for MLP on each point
        Return:
            new_points: (batch_size, ndataset1, mlp[-1]) TF tensor
    '''
    with tf.variable_scope(scope) as sc:
        dist, idx = three_nn(xyz1, xyz2)
        dist = tf.maximum(dist, 1e-10)
        norm = tf.reduce_sum((1.0/dist),axis=2,keep_dims=True)
        norm = tf.tile(norm,[1,1,3])
        weight = (1.0/dist) / norm
        interpolated_points = three_interpolate(points2, idx, weight)

        if points1 is not None:
            new_points1 = tf.concat(axis=2, values=[interpolated_points, points1]) # B,ndataset1,nchannel1+nchannel2
        else:
            new_points1 = interpolated_points
        new_points1 = tf.expand_dims(new_points1, 2)
        for i, num_out_channel in enumerate(mlp):
            if i == len(mlp)-1 and not(last_mlp_activation):
                activation_fn = None
            else:
                activation_fn = tf.nn.relu
            new_points1 = tf_util.conv2d(new_points1, num_out_channel, [1,1],
                                         padding='VALID', stride=[1,1],
                                         bn=bn, is_training=is_training,
                                         scope='conv_%d'%(i), bn_decay=bn_decay, activation_fn=activation_fn)
        new_points1 = tf.squeeze(new_points1, [2]) # B,ndataset1,mlp[-1]
        return new_points1 

def weight_net_hidden(xyz, hidden_units, scope, is_training, bn_decay=None, weight_decay=None,
                      activation_fn=tf.nn.relu, is_dist=False):
    with tf.variable_scope(scope) as sc:
        net = xyz
        for i, num_hidden_units in enumerate(hidden_units):
            net = tf_util.conv2d(net, num_hidden_units, [1, 1],
                                 padding='VALID', stride=[1, 1],
                                 bn=True, is_training=is_training, activation_fn=activation_fn,
                                 scope='wconv{}'.format(i), bn_decay=bn_decay,
                                 weight_decay=weight_decay, is_dist=is_dist)

            # net = tf_util.dropout(net, keep_prob=0.5, is_training=is_training, scope='wconv_dp{}'.format(i))
    return net 

def get_model(source_point_cloud, template_point_cloud, is_training, bn_decay=None):
	point_cloud = tf.concat([source_point_cloud, template_point_cloud],0)
	batch_size = point_cloud.get_shape()[0].value
	num_point = point_cloud.get_shape()[1].value
	end_points = {}

	input_image = tf.expand_dims(point_cloud, -1)

	net = tf_util.conv2d(input_image, 64, [1,3],
						 padding='VALID', stride=[1,1],
						 bn=False, is_training=is_training,
						 scope='conv1', bn_decay=bn_decay)
	net = tf_util.conv2d(net, 64, [1,1],
						 padding='VALID', stride=[1,1],
						 bn=False, is_training=is_training,
						 scope='conv2', bn_decay=bn_decay)
	net = tf_util.conv2d(net, 64, [1,1],
						 padding='VALID', stride=[1,1],
						 bn=False, is_training=is_training,
						 scope='conv3', bn_decay=bn_decay)
	net = tf_util.conv2d(net, 128, [1,1],
						 padding='VALID', stride=[1,1],
						 bn=False, is_training=is_training,
						 scope='conv4', bn_decay=bn_decay)
	net = tf_util.conv2d(net, 1024, [1,1],
						 padding='VALID', stride=[1,1],
						 bn=False, is_training=is_training,
						 scope='conv5', bn_decay=bn_decay)

	# Symmetric function: max pooling
	net = tf_util.max_pool2d(net, [num_point,1],
							 padding='VALID', scope='maxpool')
	net = tf.reshape(net, [batch_size, -1])
	source_global_feature = tf.slice(net, [0,0], [int(batch_size/2),1024])
	template_global_feature = tf.slice(net, [int(batch_size/2),0], [int(batch_size/2),1024])
	return source_global_feature, template_global_feature 

def get_transform(point_cloud, is_training, bn_decay=None, K = 3):
    """ Transform Net, input is BxNx3 gray image
        Return:
            Transformation matrix of size 3xK """
    batch_size = point_cloud.get_shape()[0].value
    num_point = point_cloud.get_shape()[1].value

    input_image = tf.expand_dims(point_cloud, -1)
    net = tf_util.conv2d(input_image, 64, [1,3], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='tconv1', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 128, [1,1], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='tconv3', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 1024, [1,1], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='tconv4', bn_decay=bn_decay)
    net = tf_util.max_pool2d(net, [num_point,1], padding='VALID', scope='tmaxpool')

    net = tf.reshape(net, [batch_size, -1])
    net = tf_util.fully_connected(net, 128, bn=True, is_training=is_training, scope='tfc1', bn_decay=bn_decay)
    net = tf_util.fully_connected(net, 128, bn=True, is_training=is_training, scope='tfc2', bn_decay=bn_decay)

    with tf.variable_scope('transform_XYZ') as sc:
        assert(K==3)
        weights = tf.get_variable('weights', [128, 3*K], initializer=tf.constant_initializer(0.0), dtype=tf.float32)
        biases = tf.get_variable('biases', [3*K], initializer=tf.constant_initializer(0.0), dtype=tf.float32) + tf.constant([1,0,0,0,1,0,0,0,1], dtype=tf.float32)
        transform = tf.matmul(net, weights)
        transform = tf.nn.bias_add(transform, biases)

    #transform = tf_util.fully_connected(net, 3*K, activation_fn=None, scope='tfc3')
    transform = tf.reshape(transform, [batch_size, 3, K])
    return transform 

def get_model(point_cloud, is_training, bn_decay=None):
    """ ConvNet baseline, input is BxNx3 gray image """
    batch_size = point_cloud.get_shape()[0].value
    num_point = point_cloud.get_shape()[1].value

    input_image = tf.expand_dims(point_cloud, -1)
    # CONV
    net = tf_util.conv2d(input_image, 64, [1,9], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='conv1', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 64, [1,1], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='conv2', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 64, [1,1], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='conv3', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 128, [1,1], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='conv4', bn_decay=bn_decay)
    points_feat1 = tf_util.conv2d(net, 1024, [1,1], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='conv5', bn_decay=bn_decay)
    # MAX
    pc_feat1 = tf_util.max_pool2d(points_feat1, [num_point,1], padding='VALID', scope='maxpool1')
    # FC
    pc_feat1 = tf.reshape(pc_feat1, [batch_size, -1])
    pc_feat1 = tf_util.fully_connected(pc_feat1, 256, bn=True, is_training=is_training, scope='fc1', bn_decay=bn_decay)
    pc_feat1 = tf_util.fully_connected(pc_feat1, 128, bn=True, is_training=is_training, scope='fc2', bn_decay=bn_decay)
    print(pc_feat1)
   
    # CONCAT 
    pc_feat1_expand = tf.tile(tf.reshape(pc_feat1, [batch_size, 1, 1, -1]), [1, num_point, 1, 1])
    points_feat1_concat = tf.concat(axis=3, values=[points_feat1, pc_feat1_expand])
    
    # CONV 
    net = tf_util.conv2d(points_feat1_concat, 512, [1,1], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='conv6')
    net = tf_util.conv2d(net, 256, [1,1], padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training, scope='conv7')
    net = tf_util.dropout(net, keep_prob=0.7, is_training=is_training, scope='dp1')
    net = tf_util.conv2d(net, 13, [1,1], padding='VALID', stride=[1,1],
                         activation_fn=None, scope='conv8')
    net = tf.squeeze(net, [2])

    return net 

def feature_transform_net(inputs, is_training, bn_decay=None, K=64):
    """ Feature Transform Net, input is BxNx1xK
        Return:
            Transformation matrix of size KxK """
    batch_size = inputs.get_shape()[0].value
    num_point = inputs.get_shape()[1].value

    net = tf_util.conv2d(inputs, 64, [1,1],
                         padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training,
                         scope='tconv1', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 128, [1,1],
                         padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training,
                         scope='tconv2', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 1024, [1,1],
                         padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training,
                         scope='tconv3', bn_decay=bn_decay)
    net = tf_util.max_pool2d(net, [num_point,1],
                             padding='VALID', scope='tmaxpool')

    net = tf.reshape(net, [batch_size, -1])
    net = tf_util.fully_connected(net, 512, bn=True, is_training=is_training,
                                  scope='tfc1', bn_decay=bn_decay)
    net = tf_util.fully_connected(net, 256, bn=True, is_training=is_training,
                                  scope='tfc2', bn_decay=bn_decay)

    with tf.variable_scope('transform_feat') as sc:
        weights = tf.get_variable('weights', [256, K*K],
                                  initializer=tf.constant_initializer(0.0),
                                  dtype=tf.float32)
        biases = tf.get_variable('biases', [K*K],
                                 initializer=tf.constant_initializer(0.0),
                                 dtype=tf.float32)
        biases += tf.constant(np.eye(K).flatten(), dtype=tf.float32)
        transform = tf.matmul(net, weights)
        transform = tf.nn.bias_add(transform, biases)

    transform = tf.reshape(transform, [batch_size, K, K])
    return transform 

def pointnet_fp_module(xyz1, xyz2, points1, points2, mlp, is_training, bn_decay, scope, bn=True):
    ''' PointNet Feature Propogation (FP) Module
        Input:                                                                                                      
            xyz1: (batch_size, ndataset1, 3) TF tensor                                                              
            xyz2: (batch_size, ndataset2, 3) TF tensor, sparser than xyz1                                           
            points1: (batch_size, ndataset1, nchannel1) TF tensor                                                   
            points2: (batch_size, ndataset2, nchannel2) TF tensor
            mlp: list of int32 -- output size for MLP on each point                                                 
        Return:
            new_points: (batch_size, ndataset1, mlp[-1]) TF tensor
    '''
    with tf.variable_scope(scope) as sc:
        dist, idx = three_nn(xyz1, xyz2)
        dist = tf.maximum(dist, 1e-10)
        norm = tf.reduce_sum((1.0/dist),axis=2,keep_dims=True)
        norm = tf.tile(norm,[1,1,3])
        weight = (1.0/dist) / norm
        interpolated_points = three_interpolate(points2, idx, weight)

        if points1 is not None:
            new_points1 = tf.concat(axis=2, values=[interpolated_points, points1]) # B,ndataset1,nchannel1+nchannel2
        else:
            new_points1 = interpolated_points
        new_points1 = tf.expand_dims(new_points1, 2)
        for i, num_out_channel in enumerate(mlp):
            new_points1 = tf_util.conv2d(new_points1, num_out_channel, [1,1],
                                         padding='VALID', stride=[1,1],
                                         bn=bn, is_training=is_training,
                                         scope='conv_%d'%(i), bn_decay=bn_decay)
        new_points1 = tf.squeeze(new_points1, [2]) # B,ndataset1,mlp[-1]
        return new_points1 

def feature_transform_net(inputs, is_training, bn_decay=None, K=64):
    """ Feature Transform Net, input is BxNx1xK
        Return:
            Transformation matrix of size KxK """
    batch_size = inputs.get_shape()[0].value
    num_point = inputs.get_shape()[1].value

    net = tf_util.conv2d(inputs, 64, [1,1],
                         padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training,
                         scope='tconv1', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 128, [1,1],
                         padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training,
                         scope='tconv2', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 1024, [1,1],
                         padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training,
                         scope='tconv3', bn_decay=bn_decay)
    net = tf_util.max_pool2d(net, [num_point,1],
                             padding='VALID', scope='tmaxpool')

    net = tf.reshape(net, [batch_size, -1])
    net = tf_util.fully_connected(net, 512, bn=True, is_training=is_training,
                                  scope='tfc1', bn_decay=bn_decay)
    net = tf_util.fully_connected(net, 256, bn=True, is_training=is_training,
                                  scope='tfc2', bn_decay=bn_decay)

    with tf.variable_scope('transform_feat') as sc:
        weights = tf.get_variable('weights', [256, K*K],
                                  initializer=tf.constant_initializer(0.0),
                                  dtype=tf.float32)
        biases = tf.get_variable('biases', [K*K],
                                 initializer=tf.constant_initializer(0.0),
                                 dtype=tf.float32)
        biases += tf.constant(np.eye(K).flatten(), dtype=tf.float32)
        transform = tf.matmul(net, weights)
        transform = tf.nn.bias_add(transform, biases)

    transform = tf.reshape(transform, [batch_size, K, K])
    return transform 

def pointnet_fp_module(xyz1, xyz2, points1, points2, mlp, is_training, bn_decay, scope, bn=True):
    ''' PointNet Feature Propogation (FP) Module
        Input:                                                                                                      
            xyz1: (batch_size, ndataset1, 3) TF tensor                                                              
            xyz2: (batch_size, ndataset2, 3) TF tensor, sparser than xyz1                                           
            points1: (batch_size, ndataset1, nchannel1) TF tensor                                                   
            points2: (batch_size, ndataset2, nchannel2) TF tensor
            mlp: list of int32 -- output size for MLP on each point                                                 
        Return:
            new_points: (batch_size, ndataset1, mlp[-1]) TF tensor
    '''
    with tf.variable_scope(scope) as sc:
        dist, idx = three_nn(xyz1, xyz2)
        dist = tf.maximum(dist, 1e-10)
        norm = tf.reduce_sum((1.0/dist),axis=2,keep_dims=True)
        norm = tf.tile(norm,[1,1,3])
        weight = (1.0/dist) / norm
        interpolated_points = three_interpolate(points2, idx, weight)

        if points1 is not None:
            new_points1 = tf.concat(axis=2, values=[interpolated_points, points1]) # B,ndataset1,nchannel1+nchannel2
        else:
            new_points1 = interpolated_points
        new_points1 = tf.expand_dims(new_points1, 2)
        for i, num_out_channel in enumerate(mlp):
            new_points1 = tf_util.conv2d(new_points1, num_out_channel, [1,1],
                                         padding='VALID', stride=[1,1],
                                         bn=bn, is_training=is_training,
                                         scope='conv_%d'%(i), bn_decay=bn_decay)
        new_points1 = tf.squeeze(new_points1, [2]) # B,ndataset1,mlp[-1]
        return new_points1 

def get_model(source_point_cloud, template_point_cloud, is_training, bn_decay=None):
	point_cloud = tf.concat([source_point_cloud, template_point_cloud],0)
	batch_size = point_cloud.get_shape()[0].value
	num_point = point_cloud.get_shape()[1].value
	end_points = {}

	input_image = tf.expand_dims(point_cloud, -1)

	net = tf_util.conv2d(input_image, 64, [1,3],
						 padding='VALID', stride=[1,1],
						 bn=False, is_training=is_training,
						 scope='conv1', bn_decay=bn_decay)
	net = tf_util.conv2d(net, 64, [1,1],
						 padding='VALID', stride=[1,1],
						 bn=False, is_training=is_training,
						 scope='conv2', bn_decay=bn_decay)
	net = tf_util.conv2d(net, 64, [1,1],
						 padding='VALID', stride=[1,1],
						 bn=False, is_training=is_training,
						 scope='conv3', bn_decay=bn_decay)
	net = tf_util.conv2d(net, 128, [1,1],
						 padding='VALID', stride=[1,1],
						 bn=False, is_training=is_training,
						 scope='conv4', bn_decay=bn_decay)
	net = tf_util.conv2d(net, 1024, [1,1],
						 padding='VALID', stride=[1,1],
						 bn=False, is_training=is_training,
						 scope='conv5', bn_decay=bn_decay)

	# Symmetric function: max pooling
	net = tf_util.max_pool2d(net, [num_point,1],
							 padding='VALID', scope='maxpool')
	net = tf.reshape(net, [batch_size, -1])
	source_global_feature = tf.slice(net, [0,0], [int(batch_size/2),1024])
	template_global_feature = tf.slice(net, [int(batch_size/2),0], [int(batch_size/2),1024])
	return source_global_feature, template_global_feature 

def get_center_regression_net(object_point_cloud, one_hot_vec,
                              is_training, bn_decay, end_points):
    ''' Regression network for center delta. a.k.a. T-Net.
    Input:
        object_point_cloud: TF tensor in shape (B,M,C)
            point clouds in 3D mask coordinate
        one_hot_vec: TF tensor in shape (B,3)
            length-3 vectors indicating predicted object type
    Output:
        predicted_center: TF tensor in shape (B,3)
    ''' 
    num_point = object_point_cloud.get_shape()[1].value
    net = tf.expand_dims(object_point_cloud, 2)
    net = tf_util.conv2d(net, 128, [1,1],
                         padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training,
                         scope='conv-reg1-stage1', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 128, [1,1],
                         padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training,
                         scope='conv-reg2-stage1', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 256, [1,1],
                         padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training,
                         scope='conv-reg3-stage1', bn_decay=bn_decay)
    net = tf_util.max_pool2d(net, [num_point,1],
        padding='VALID', scope='maxpool-stage1')
    net = tf.squeeze(net, axis=[1,2])
    net = tf.concat([net, one_hot_vec], axis=1)
    net = tf_util.fully_connected(net, 256, scope='fc1-stage1', bn=True,
        is_training=is_training, bn_decay=bn_decay)
    net = tf_util.fully_connected(net, 128, scope='fc2-stage1', bn=True,
        is_training=is_training, bn_decay=bn_decay)
    predicted_center = tf_util.fully_connected(net, 3, activation_fn=None,
        scope='fc3-stage1')
    return predicted_center, end_points 

def pointnet_fp_module(xyz1, xyz2, points1, points2, mlp, is_training, bn_decay, scope, bn=True):
    ''' PointNet Feature Propogation (FP) Module
        Input:                                                                                                      
            xyz1: (batch_size, ndataset1, 3) TF tensor                                                              
            xyz2: (batch_size, ndataset2, 3) TF tensor, sparser than xyz1                                           
            points1: (batch_size, ndataset1, nchannel1) TF tensor                                                   
            points2: (batch_size, ndataset2, nchannel2) TF tensor
            mlp: list of int32 -- output size for MLP on each point                                                 
        Return:
            new_points: (batch_size, ndataset1, mlp[-1]) TF tensor
    '''
    with tf.variable_scope(scope) as sc:
        dist, idx = three_nn(xyz1, xyz2)
        dist = tf.maximum(dist, 1e-10)
        norm = tf.reduce_sum((1.0/dist),axis=2,keep_dims=True)
        norm = tf.tile(norm,[1,1,3])
        weight = (1.0/dist) / norm
        interpolated_points = three_interpolate(points2, idx, weight)

        if points1 is not None:
            new_points1 = tf.concat(axis=2, values=[interpolated_points, points1]) # B,ndataset1,nchannel1+nchannel2
        else:
            new_points1 = interpolated_points
        new_points1 = tf.expand_dims(new_points1, 2)
        for i, num_out_channel in enumerate(mlp):
            new_points1 = tf_util.conv2d(new_points1, num_out_channel, [1,1],
                                         padding='VALID', stride=[1,1],
                                         bn=bn, is_training=is_training,
                                         scope='conv_%d'%(i), bn_decay=bn_decay)
        new_points1 = tf.squeeze(new_points1, [2]) # B,ndataset1,mlp[-1]
        return new_points1 

def pointnet_sa_module_msg(xyz, points, npoint, radius_list, nsample_list, mlp_list, is_training, bn_decay, scope, bn=True, use_xyz=True, use_nchw=False):
    ''' PointNet Set Abstraction (SA) module with Multi-Scale Grouping (MSG)
        Input:
            xyz: (batch_size, ndataset, 3) TF tensor
            points: (batch_size, ndataset, channel) TF tensor
            npoint: int32 -- #points sampled in farthest point sampling
            radius: list of float32 -- search radius in local region
            nsample: list of int32 -- how many points in each local region
            mlp: list of list of int32 -- output size for MLP on each point
            use_xyz: bool, if True concat XYZ with local point features, otherwise just use point features
            use_nchw: bool, if True, use NCHW data format for conv2d, which is usually faster than NHWC format
        Return:
            new_xyz: (batch_size, npoint, 3) TF tensor
            new_points: (batch_size, npoint, \sum_k{mlp[k][-1]}) TF tensor
    '''
    data_format = 'NCHW' if use_nchw else 'NHWC'
    with tf.variable_scope(scope) as sc:
        new_xyz = gather_point(xyz, farthest_point_sample(npoint, xyz))
        new_points_list = []
        for i in range(len(radius_list)):
            radius = radius_list[i]
            nsample = nsample_list[i]
            idx, pts_cnt = query_ball_point(radius, nsample, xyz, new_xyz)
            grouped_xyz = group_point(xyz, idx)
            grouped_xyz -= tf.tile(tf.expand_dims(new_xyz, 2), [1,1,nsample,1])
            if points is not None:
                grouped_points = group_point(points, idx)
                if use_xyz:
                    grouped_points = tf.concat([grouped_points, grouped_xyz], axis=-1)
            else:
                grouped_points = grouped_xyz
            if use_nchw: grouped_points = tf.transpose(grouped_points, [0,3,1,2])
            for j,num_out_channel in enumerate(mlp_list[i]):
                grouped_points = tf_util.conv2d(grouped_points, num_out_channel, [1,1],
                                                padding='VALID', stride=[1,1], bn=bn, is_training=is_training,
                                                scope='conv%d_%d'%(i,j), bn_decay=bn_decay)
            if use_nchw: grouped_points = tf.transpose(grouped_points, [0,2,3,1])
            new_points = tf.reduce_max(grouped_points, axis=[2])
            new_points_list.append(new_points)
        new_points_concat = tf.concat(new_points_list, axis=-1)
        return new_xyz, new_points_concat 

def get_center_regression_net(object_point_cloud, one_hot_vec,
                              is_training, bn_decay, end_points):
    ''' Regression network for center delta. a.k.a. T-Net.
    Input:
        object_point_cloud: TF tensor in shape (B,M,C)
            point clouds in 3D mask coordinate
        one_hot_vec: TF tensor in shape (B,3)
            length-3 vectors indicating predicted object type
    Output:
        predicted_center: TF tensor in shape (B,3)
    ''' 
    num_point = object_point_cloud.get_shape()[1].value#M
    net = tf.expand_dims(object_point_cloud, 2)#(B,M,1,C)
    net = tf_util.conv2d(net, 128, [1,1],
                         padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training,
                         scope='conv-reg1-stage1', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 128, [1,1],
                         padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training,
                         scope='conv-reg2-stage1', bn_decay=bn_decay)
    net = tf_util.conv2d(net, 256, [1,1],
                         padding='VALID', stride=[1,1],
                         bn=True, is_training=is_training,
                         scope='conv-reg3-stage1', bn_decay=bn_decay)
    net = tf_util.max_pool2d(net, [num_point,1],
        padding='VALID', scope='maxpool-stage1')
    net = tf.squeeze(net, axis=[1,2])
    net = tf.concat([net, one_hot_vec], axis=1)
    net = tf_util.fully_connected(net, 256, scope='fc1-stage1', bn=True,
        is_training=is_training, bn_decay=bn_decay)
    net = tf_util.fully_connected(net, 128, scope='fc2-stage1', bn=True,
        is_training=is_training, bn_decay=bn_decay)
    predicted_center = tf_util.fully_connected(net, 3, activation_fn=None,
        scope='fc3-stage1')
    return predicted_center, end_points