issues
search
craston
/
MARS
MARS: Motion-Augmented RGB Stream for Action Recognition
MIT License
161
stars
45
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
there was a typo while accessing an attribute
#54
VanIro
closed
1 year ago
0
UCF 101 Training Guide
#53
sepehratwork
opened
2 years ago
0
about Accuracy
#52
wutongshu819274135
opened
2 years ago
0
about flow_graphs
#51
wutongshu819274135
opened
2 years ago
0
a bug in code when train/test ucf101
#50
Maojianzeng
opened
2 years ago
0
RuntimeError: Function AddmmBackward returned an invalid gradient at index 1 - got [1, 2048] but expected shape compatible with [1, 32768]?
#49
Tariq195
opened
2 years ago
0
Please tell me about your optical flow directory, so that I can format it the way your code accept?
#48
Tariq195
opened
3 years ago
0
Can you tell me the shape of your input flow?
#47
Tariq195
opened
3 years ago
0
Code error: Normalize code does not apply
#46
seominseok0429
opened
3 years ago
0
How to train MERS model?
#45
HuangZuShu
opened
3 years ago
0
Can you tell me, what was your input size to the model? just tell me for single batch_size?
#44
Tariq195
closed
3 years ago
1
How can I give input to 3d CNN?
#43
Tariq195
closed
3 years ago
3
For Flow stream use test_single_stream.py
#42
Maojianzeng
closed
3 years ago
4
how to train two-stream(RGB+Flow) Network?
#41
XDLiuyyy
closed
3 years ago
1
Flow accuracy
#40
owen199711
opened
3 years ago
1
How to accelerate the extraction of optical flow ?
#39
HuangZuShu
opened
3 years ago
3
Whether there is a bug in resnext.py ?
#38
HuangZuShu
opened
3 years ago
0
where can I get the trained models like MARS on HMDB51-2,3
#37
runtoforever
closed
3 years ago
1
MERS weights request
#36
liamhiley
opened
3 years ago
0
the accuracy is lower than 72.2% in HMDB51-1
#35
runtoforever
closed
3 years ago
9
Update test_two_stream.py
#34
dimbyr
closed
3 years ago
1
Update test_two_stream.py
#33
dimbyr
closed
3 years ago
0
Update test_two_stream.py
#32
dimbyr
closed
3 years ago
0
The accuracy of validation is low
#31
hushunda
opened
3 years ago
1
The data set code seems to be out of date (and likely not functioning)
#30
searchivarius
closed
3 years ago
2
HMDB51 testing error!!!!
#29
KP1-cmd
closed
3 years ago
3
how to get the label file of HMDB51
#28
wen-flow
opened
4 years ago
9
the setting of weight_decay
#27
JianhaoZhan
closed
4 years ago
1
Question about preprocessing data
#26
henryle97
closed
4 years ago
1
Could you share the MiniKinetics trained model?
#25
loatlam
closed
4 years ago
1
The accuracy of validation during training RGB stream is basically unchanged
#24
Abby-Wu
opened
4 years ago
1
Running SomeThingSomeThingV1 test_RGB problem.
#23
loatlam
opened
4 years ago
1
Run pre-trained model on single video
#22
pittyacg
closed
3 years ago
1
extract flow problem
#21
cym3223
opened
4 years ago
15
Flow Network For Fine-Tuning
#20
artest08
closed
3 years ago
1
Could you show some examples about how to train Flow model on something-v1?
#19
Zhuysheng
closed
4 years ago
1
Could you share all of the finetune hyperparameters on hmdb51?
#18
seominseok0429
closed
3 years ago
1
The test script does not work.
#17
seominseok0429
closed
4 years ago
1
Confusion Matrix script for given datasets.
#16
bytecode-Ninja
closed
4 years ago
0
The speed to extract the flow map is too slow
#15
GuardSkill
closed
3 years ago
5
A small query about finetuning
#14
DAVEISHAN
closed
4 years ago
2
Class Activation Map visualizing script.
#13
bytecode-Ninja
closed
4 years ago
0
About MRS+Flow+RGB testing
#12
errllxj
closed
4 years ago
12
how to test three-stream result for rgb+mars+flow
#11
zwx980624
closed
4 years ago
1
One more question about training the flow_HMDB51_64f model
#10
qiuqicao77
closed
4 years ago
2
Question about training 64f parameter?
#9
qiuqicao77
closed
4 years ago
2
HMDB51 training error!!!!
#8
qiuqicao77
closed
4 years ago
3
Training UCF-101 dataset
#7
Fazlik995
opened
4 years ago
1
UCF101 dataset train error
#6
Fazlik995
opened
5 years ago
10
percentage difference for previous SOTA paper
#5
Fazlik995
closed
5 years ago
2
Next