TensorFlow中对3D卷积的扩展支持

时间:2018-12-07 16:02:26

标签: tensorflow deep-learning conv-neural-network

我正在尝试在视频上运行Dilated 3D卷积,但是反复出现错误

  

NotFoundError(请参阅上面的回溯):没有算法可用!        [[节点Conv3D_1(定义为:7)= Conv3D [T = DT_FLOAT,data_format =“ NDHWC”,膨胀= [1,2,2,2,1],padding =“ SAME”,步幅= [1,1, 1,1,1],_device =“ / job:localhost /副本:0 /任务:0 / device:GPU:0”](_ arg_Placeholder_1_0_0 / _1,Variable_1 / read)]]        [[{{node Conv3D_1 / _3}} = _Recvclient_terminated = false,recv_device =“ / job:localhost /副本:0 / task:0 /设备:CPU:0”,send_device =“ / job:localhost /副本:0 / task:0 / device:GPU:0“,send_device_incarnation = 1,tensor_name =” edge_8_Conv3D_1“,tensor_type = DT_FLOAT,_device =” / job:localhost /副本:0 / task:0 / device:CPU:0“]] < / p>

我不明白问题是什么? TIA

0 个答案:

没有答案