issues
search
m-bain
/
frozen-in-time
Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval [ICCV'21]
https://arxiv.org/abs/2104.00650
MIT License
351
stars
44
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Can't experiment with other ViT backbones
#60
minhhoang2705
closed
2 months ago
0
Cannot reproduce results
#59
mgwillia
opened
1 year ago
0
RuntimeError: There were no tensor arguments to this function (e.g., you passed an empty list of Tensors), but no fallback function is registered for schema aten::_cat.
#58
fishfree
opened
1 year ago
1
Could you please also share the demo UI codes?
#57
fishfree
opened
1 year ago
2
Question about pretrained models?
#56
valeriomieuli
closed
2 years ago
1
pre-process mp4
#55
nattikahana
closed
2 years ago
1
About the effects of sliding_window_stride
#54
xiangyh9988
opened
2 years ago
0
Result about MSVD
#53
shufangxun
closed
2 years ago
2
CC3M dataset broken
#52
shufangxun
closed
2 years ago
1
Code/template for the demo?
#51
scottfleming
closed
2 years ago
4
config about Paper Result
#50
shufangxun
closed
2 years ago
2
How to select checkpoint when fine-tuning on MSRVTT
#49
chaochen99
closed
2 years ago
1
evaluation for MSVD .
#48
chaochen99
closed
2 years ago
1
Creating 256 emb vector from text during inferenece
#47
yulies
closed
2 years ago
2
How about initializing parameters with CLIP model
#46
realTaki
closed
2 years ago
1
Curriculum Learning and Video-Image Joint Training
#45
vateye
closed
2 years ago
4
License ?
#44
zvr
closed
2 years ago
2
What does zero-pad means?
#43
Andy1621
closed
2 years ago
1
About Curriculum Learning
#42
Andy1621
closed
2 years ago
3
Which results in paper correspond to the finetune command?
#41
akira-l
closed
2 years ago
5
CC3M data error
#40
akira-l
closed
2 years ago
3
Can you share some recordings of your experiments
#39
realTaki
opened
2 years ago
2
How can I use the pretrain results?
#38
Qi-Zhangyang
closed
2 years ago
3
Hi, can you upload the dataset file for MSVD and DiDeMo?
#37
justopit
opened
2 years ago
0
Reproducibility
#36
zixuwang1996
closed
2 years ago
1
why noy evaluate your model on HowTo100 dataset
#35
ForawardStar
closed
2 years ago
2
Looking for a simple example, how to use, like CLIP
#34
lonngxiang
opened
2 years ago
3
Broken links
#33
lucas-ventura
closed
3 years ago
1
Can you provide the checkpoint that was only trained on WebVid?
#32
bryant1410
closed
3 years ago
1
Fine tune on a custom dataset
#31
avinashsai
closed
3 years ago
1
Frame sampling in test phase
#30
YizhaoGao-AI
closed
3 years ago
2
The pre-training budget on WebVid-2M
#29
mzhaoshuai
closed
3 years ago
1
Test set of MSR-VTT for downstream evaluation
#28
geyuying
closed
3 years ago
5
Hi, could you please provide a zip file for DiDeMo dataset?
#27
NoelVouitsis
opened
3 years ago
0
Fix `item()` called multiple times
#26
bryant1410
closed
3 years ago
0
Move env to environment.yml
#25
bryant1410
closed
3 years ago
0
Use tqdm for training and validation
#24
bryant1410
closed
3 years ago
1
Avoid using DataParallel for some batch sizes
#23
bryant1410
closed
3 years ago
0
Off-by-one issues with the frame sampling
#22
bryant1410
opened
3 years ago
0
Set Decord as default reader in data loader
#21
bryant1410
closed
3 years ago
0
Revert "Add missing torch.nn import in trainer.py"
#20
m-bain
closed
3 years ago
0
Add missing torch.nn import in trainer.py
#19
bryant1410
closed
3 years ago
0
Assert video file exists
#18
bryant1410
closed
3 years ago
0
Add support for Decord video reader
#17
bryant1410
closed
3 years ago
1
Bad file descriptor
#16
bryant1410
closed
3 years ago
2
Finetuning the pretrained model on MSR-VTT
#15
geyuying
closed
3 years ago
3
Fix val loss minimization in MSR-VTT config file
#14
bryant1410
closed
3 years ago
0
Watermarks influence
#13
FuTSy13
closed
3 years ago
1
What to run to reproduce the best results?
#12
bryant1410
closed
3 years ago
2
Config for the method trained on CC3M
#11
bryant1410
closed
3 years ago
3
Next