training PVA-lite model

sanghoon / pva-faster-rcnn

Demo code for PVANet

https://arxiv.org/abs/1611.08588

Other

650 stars 240 forks source link

training PVA-lite model #74

Open adityasaha opened 7 years ago

adityasaha commented 7 years ago

Can you help me to run PVA-lite model with sample command line arguments for PASCAL VOC 2007 dataset

adityasaha commented 7 years ago

@sanghoon

yzhang123 commented 7 years ago

python tools/train_net.py --gpu 0 --solver models/pvanet/example_train/solver.prototxt --weight models/pvanet/pretrained/pva9.1_pretrained_no_fc6.caffemodel --iters 100000 --cfg models/pvanet/cfgs/train.yml --imdb voc_2007_trainval

adityasaha commented 7 years ago

Thanks a lot for your reply! I have also found in the paper the PVANET+Compressed as better FPS. Can you please tell me how to train and test a compressed model. If you can provide command line arguments for the PVANET+compressed model that will be great! Again I appreciate your reply to my previous queries!

adityasaha commented 7 years ago

@yzhang123

adityasaha commented 7 years ago

@sanghoon

twmht commented 6 years ago

@adityasaha

where you find PVA-lite? which paper?