Visualization for comparison

[nikky4D]

Hi, I would like to test my implementation to make sure it matches up with yours. Would you have a test video or videos you recommend I use to compare my results to yours? Just to make sure it is running properly.

Also, if I wanted to visualize the features, is there a code I can use for that?

Thank you for all your help.

[LuoweiZhou]

@nikky4D We have a code to visualize the BBox in GVD, but not for features.

I'd suggest you simply print out the sum or norm of the feature file and compare it against our provided file. You can also do it at a finer level, say indexing a block from the tensor to see if the values match our file, like I mentioned here: https://github.com/facebookresearch/grounded-video-description/issues/20#issuecomment-610036948 If you see error around or < 1e-5, this is normal due to precision errors.

[nikky4D]

Thank you. I'll get to work on it over the next two weeks. Thank you for all your help.

[MarcusNerva]

@nikky4D Hi amigo! Which enviroment did you choose for detectron-vlp ? GVD or VLP? Or just prepare conda environment by yourself? Would you please share your requirements.txt?

[nikky4D]

@MarcusNerva I was working in colab, and using the install process for detectron-vlp. My environment was based on GVD but had to be modified to get it working in colab.

[MarcusNerva]

@nikky4D Could you tell me the version of choosed pytorch? I meet this problem:

[I net_async_base.h:205] Using specified CPU pool size: 4; device id: -1 [I net_async_base.h:210] Created new CPU pool, size: 4; device id: -1 [E net_async_base.cc:377] [enforce fail at math_gpu.cu:569] status == CUBLAS_STATUS_SUCCESS. 13 vs 0. Error at: /opt/conda/conda-bld/pytorch_1556653000816/work/caffe2/utils/math_gpu.cu:569: CUBLAS_STATUS_EXECUTION_FAILED Error from operator: input: "gpu_0/fc7" input: "gpu_0/bbox_pred_w" input: "gpu_0/bbox_pred_b" output: "gpu_0/bbox_pred" name: "" type: "FC" arg { name: "use_cudnn" i: 1 } arg { name: "cudnn_exhaustive_search" i: 0 } arg { name: "order" s: "NCHW" } device_option { device_type: 1 device_id: 0 }frame #0: c10::ThrowEnforceNotMet(char const, int, char const, std::string const&, void const) + 0x59 (0x7f65a9797409 in /home/marcusnerva/anaconda3/envs/gvd_pt1_1/lib/python2.7/site-packages/caffe2/python/../../torch/lib/libc10.so) frame #1: void caffe2::math::Gemm<float, caffe2::CUDAContext, caffe2::DefaultEngine>(CBLAS_TRANSPOSE, CBLAS_TRANSPOSE, int, int, int, float, float const, float const, float, float, caffe2::CUDAContext*, caffe2::TensorProto_DataType) + 0x6da (0x7f65ac6f399a in /home/marcusnerva/anaconda3/envs/gvd_pt1_1/lib/python2.7/site-packages/caffe2/python/../../torch/lib/libcaffe2_gpu.so) frame #2: + 0x14e3730 (0x7f65ab326730 in /home/marcusnerva/anaconda3/envs/gvd_pt1_1/lib/python2.7/site-packages/caffe2/python/../../torch/lib/libcaffe2_gpu.so) frame #3: + 0x14d8e88 (0x7f65ab31be88 in /home/marcusnerva/anaconda3/envs/gvd_pt1_1/lib/python2.7/site-packages/caffe2/python/../../torch/lib/libcaffe2_gpu.so) frame #4: + 0x13cb0b5 (0x7f65ab20e0b5 in /home/marcusnerva/anaconda3/envs/gvd_pt1_1/lib/python2.7/site-packages/caffe2/python/../../torch/lib/libcaffe2_gpu.so) frame #5: caffe2::AsyncNetBase::run(int, int) + 0x144 (0x7f65cdd1cb94 in /home/marcusnerva/anaconda3/envs/gvd_pt1_1/lib/python2.7/site-packages/caffe2/python/../../torch/lib/libcaffe2.so) frame #6: + 0x168f009 (0x7f65cdd23009 in /home/marcusnerva/anaconda3/envs/gvd_pt1_1/lib/python2.7/site-packages/caffe2/python/../../torch/lib/libcaffe2.so) frame #7: c10::ThreadPool::main_loop(unsigned long) + 0x2a3 (0x7f65a97912f3 in /home/marcusnerva/anaconda3/envs/gvd_pt1_1/lib/python2.7/site-packages/caffe2/python/../../torch/lib/libc10.so) frame #8: + 0xb8678 (0x7f65e4a5b678 in /home/marcusnerva/anaconda3/envs/gvd_pt1_1/bin/../lib/libstdc++.so.6) frame #9: + 0x9609 (0x7f65eb05d609 in /lib/x86_64-linux-gnu/libpthread.so.0) frame #10: clone + 0x43 (0x7f65eae2a293 in /lib/x86_64-linux-gnu/libc.so.6) , op FC

So confusing.