Error executing sh scripts/extract_feature_hmdb51.sh, Cannot open file: r2plus1d_8.mdl

[953585130]

INFO:video_model:Number of middle filters: 576 INFO:video_model:Number of middle filters: 576 INFO:video_model:Number of middle filters: 576 INFO:video_model:Number of middle filters: 576 INFO:video_model:Number of middle filters: 576 INFO:video_model:Number of middle filters: 576 INFO:video_model:Number of middle filters: 576 INFO:video_model:Number of middle filters: 576 INFO:video_model:Number of middle filters: 576 INFO:video_model:Number of middle filters: 576 INFO:video_model:Number of middle filters: 576 INFO:video_model:Number of middle filters: 921 INFO:video_model:Number of middle filters: 1152 INFO:video_model:Number of middle filters: 1152 INFO:video_model:Number of middle filters: 1152 INFO:video_model:Number of middle filters: 1152 INFO:video_model:Number of middle filters: 1152 INFO:data_parallel_model:Parameter update function not defined --> only forward INFO:model_helper:Loading path: r2plus1d_8.mdl Traceback (most recent call last): File "tools/extract_features.py", line 328, in main() File "tools/extract_features.py", line 323, in main ExtractFeatures(args) File "tools/extract_features.py", line 142, in ExtractFeatures model_helper.LoadModel(args.load_model_path, args.db_type) File "/home/baihao/R2Plus1D-master/tools/utils/model_helper.py", line 90, in LoadModel meta_net_def = pred_exp.load_from_db(path, dbtype) File "/usr/local/lib/python2.7/dist-packages/caffe2/python/predictor/predictor_exporter.py", line 211, in load_from_db assert workspace.RunOperatorOnce(create_db), ( File "/usr/local/lib/python2.7/dist-packages/caffe2/python/workspace.py", line 165, in RunOperatorOnce return C.run_operatoronce(StringifyProto(operator)) RuntimeError: [enforce fail at db.cc:140] file. Cannot open file: r2plus1d_8.mdl Error from operator: output: "!!PREDICTOR_DBREADER" name: "" type: "CreateDB" arg { name: "db_type" s: "minidb" } arg { name: "db" s: "r2plus1d_8.mdl" }

[dutran]

It says "Cannot open file: r2plus1d_8.mdl", have you got your model downloaded?

[953585130]

@dutran I did not find the corresponding file in the pre-training model provided in Tutorial 5. The pre-training models provided are all pkl files, no mdl files. Is there no release of r2plus1d_8.mdl or is there a problem with my environment?

for ((i=1;i<=2;i++)); \ do \ python tools/extract_features.py \ --test_data=/data/users/trandu/datasets/hmdb51_feature_extraction/hmdb51_test_01_video_id_densel32$i \ --model_name=r2plus1d --model_depth=34 --clip_length_rgb=32 \ --gpus=0,1,2,3,4,5,6,7 \ --batch_size=4 \ --load_model_path=r2plus1d_8.mdl --db_type=minidb \ --output_path=/data/users/trandu/datasets/hmdb51_features/ft/hmdb51_test_01_video_id_densel32$i.pkl \ --features=softmax,label,video_id \ --sanity_check=1 --get_video_id=1 --use_local_file=1 --num_labels=51; \ done

[dutran]

I think the models are supposed to be downloaded from here: https://github.com/facebookresearch/R2Plus1D/blob/master/tutorials/models.md

[953585130]

@dutran I have solved this problem. When I successfully fine-tune the hmde51 dataset, the r2plus1d_8.mdl file is automatically generated. This is the result: INFO:train_net:Test accuracy: 0.605

After the fine-tuning is over, I plan to follow the tutorial to achieve dense predictions. This problem occurs when I execute sh scripts/extract_feature_hmdb51.sh: [mpeg4 @ 0x7ebe90082d60] Video uses a non-standard and wasteful way to store B-frames ('packed B-frames'). Consider using the mpeg4_unpack_bframes bitstream filter Without encoding but stream copy to fix it. 10260/20677 iterations INFO:feature_extractor:10260/20677 iterations [mpeg4 @ 0x7ebe94010dc0] Video uses a non-standard and wasteful way to store B-frames ('packed B-frames'). Consider using the mpeg4_unpack_bframes bitstream filter without encoding but stream copy to fix it.

No error, it has been constantly appearing, will this have an impact on the results?

[dutran]

I expect very little impact on the results.