gsig / actor-observer Goto Github PK

The command I used is python third_to_first_person.py.
and then with some problems

1. the CharadesEgo_val_video part always 0 samples loaded?

cachefile ./caches/third_to_first_person//CharadesEgo_train.pkl
Loading cached result from './caches/third_to_first_person//CharadesEgo_train.pkl'
516965 samples loaded
cachefile ./caches/third_to_first_person//CharadesEgo_val.pkl
Loading cached result from './caches/third_to_first_person//CharadesEgo_val.pkl'
137321 samples loaded
cachefile ./caches/third_to_first_person//CharadesEgo_val_video.pkl
Loading cached result from './caches/third_to_first_person//CharadesEgo_val_video.pkl'
0 samples loaded

2. Which dataset does the code use? CharadesEgo_v1_rgb or Charades_v1_rgb?

Thank you in advance!

Hi,
how to fine-tune the model from the third-view dataset on the first-view dataset?

I tried to directly fine-tune the trained model from https://github.com/gsig/charades-algorithms by using this script https://github.com/gsig/actor-observer/blob/master/exp/baseline_resnet152imagenet.py.

However, the results is only 22, which is the same with the model fine-tuned from ImageNet, which indicates the third-view pre-training has no effect on the first-view performance.
Could you please indicate how to correctly fine-tune the model trained from the third-view dataset?

Thanks.

thank you !

Sorry to disturb you!
Should base_x be replaced with base_y?
w_y = self.third_fc(base_x).view(-1) * torch.exp(self.third_scale)

And What does this function do? (in tasks.py)
def best_one_sec_moment(mat, winsize=6):

thank you!

Hi, thank you for sharing this amazing code.
How many GPUs are you used to train the model, and what type of the GPUs?
Thank you in advance!

Hi!
What's the difference between “same person ”and "different person" in paper?

Hi Gunnar,

I was just wondering what "ResNet-152 Transfer" stands for.
The papers said "It uses the Charades model to predict the activities in the third person video,
and then uses those labels as supervision for the first-person video."

Does it mean you run the Charades model on CharadesEgo_only3rd videos, then warp the labels to their corresponding 1st person video, and use the labels as supervision to finetune the original Charades model?

Thank you.

Best,
Ziwei

Hi~
In the Charades and CharadesEgo dataset, one video always contains several actions. In the code, you divide the video into several clips according to the start and end time, but I have observed that one frame may belongs to multiple action tags.
In this case, Can the loss function be trained normally?

Hi,
I find that you use version0 under the folder datasets/labels, but when I download the CharadesEgo dataset I get the label of version1. What version do you use to get the result of the paper?
Thanks.

The command I used is python baseline_resnet152charades.py.
Thank you in advance!

=> using pre-trained model 'resnet152'
loading pretrained-weights from /nfs.yoda/gsigurds/charades_pretrained/resnet_rgb.pth.tar
Traceback (most recent call last):
File "baseline_resnet152charades.py", line 38, in
main()
File "./main.py", line 60, in main
model, criterion, optimizer = create_model(args)
File "./models/init.py", line 10, in create_model
model = load_architecture(args)
File "./models/utils.py", line 78, in load_architecture
model = generic_load(args.arch, args.pretrained, args.pretrained_weights, args)
File "./models/utils.py", line 61, in generic_load
model = model.dictarch
File "./models/ActorObserverBase.py", line 55, in init
model = load_sub_architecture(args)
File "./models/utils.py", line 73, in load_sub_architecture
model = generic_load(args.subarch, args.pretrained, args.pretrained_subweights, args)
File "./models/utils.py", line 65, in generic_load
chkpoint = torch.load(weights)
File "/home/csdept/anaconda3/envs/py27/lib/python2.7/site-packages/torch/serialization.py", line 301, in load
f = open(f, 'rb')
IOError: [Errno 2] No such file or directory: '/nfs.yoda/gsigurds/charades_pretrained/resnet_rgb.pth.tar'