repeat experment for "example_train_384"

[hengck23]

I want to verify my setup is correct or not. I intend to follow : https://github.com/sanghoon/pva-faster-rcnn/tree/master/models/pvanet/example_train_384,

Training for 100k iterations (toy) tools/train_net.py --gpu 0 --solver models/pvanet/example_train_384/solver.prototxt --weights models/pvanet/imagenet/original.model --iters 100000 --cfg models/pvanet/cfgs/train.yml --imdb voc_2007_trainval

For the 100k iterations (toy) above, what results do I get if I test on imdb voc_2007_test?

'' It is mentioned that this is not the same as the paper, so i do not have reference results for checking. '' If anyone can share the results or train log files, it would be nice. Thanks!

[hengck23]

there are some answers for this: https://github.com/sanghoon/pva-faster-rcnn/issues/7 https://github.com/sanghoon/pva-faster-rcnn/issues/10

Here are the results: @duanLH gets Mean AP is 0.7190 @xiaoxiongli gets Mean AP is 0.7140

My training is still running. If I get the same results, that would be the performance and i will close the issue.

[hengck23]

Here is my results: Mean AP is 0.7136

[learntolearn11]

I have trained the experiment_finetue, but I can not find the train log file, where is the log file,?

[AITTSMD]

hi,I wonder how you draw the picture of the result? Can u give me some hints? the log file is in experiments/logs/

[etworker]

I have repeat the training process on train_384 with VOC2007, iteration times is about 1.1M, mAP on test set is about 72%, lower than 84% in paper.

maybe the reason is below?

PVANet was trained by our in-house deep learning library, not by this implementation.