chengtao.Wang

eg. rfcn_voc-0009.params should in which folder?

AssertionError: rpn_conv_3x3_weight not initialized

1

File "D:\Workspace\Workspace2\mx-rcnn-master\symnet\model.py", line 40, in check_shape assert k in arg_params, '%s not initialized' % k AssertionError: rpn_conv_3x3_weight not initialized

(no backtrace available

1

`//加载pytorch模型 auto x = torch::jit::load(weight); module_ = std::make_shared(x);` " (initDeviceProperty at ..\\aten\\src\\ATen\\cuda\\CUDAContext.cpp:26)\n**(no backtrace available**)

'module' object is not callable

2

merge6 = merge([drop4,up6], mode = 'concat', concat_axis = 3) 'module' object is not callable unet.py line 120, in get_unet

Hi，I run FCN.py and got model.ckpt-2500.data-00000-of-00001... Is there a demo about how to use the trained model？Tks!

Can It be speedup?

1

load params: data/yolov3.weights 1 predicted in 5.306 seconds, tt1 ---------------------------------------------------------- How can I speedup? for example, only detect person, by using less time. Thank you very much!

if self.sym is not None: #in line 224 发生异常: exceptions.AttributeError 'Nesterov' object has no attribute 'sym' , -- why?