Sense-X/UniFormer

Issues

• How to custom train of unifomer for video classification on my own custom dataset

  #136 opened 2 months ago by AbrarKhan009
  0

• UniFormer-S† head_dim=64, ImageNet-1K pretrained (224x224)

  #135 opened 2 months ago by atonyo11
  0

• Feature Extraction

  #133 opened 3 months ago by JonathanFlores2503
  1

• About training on my own dataset for video classification

  #134 opened 5 months ago by ChosenOne23
  1

• Precision

  #132 opened 5 months ago by anicolson
  3

• 大佬您好，我想问一下，关于UniFormer，如果只准备了解模型架构的话，代码是不是只用看这个文件https://github.com/Sense-X/UniFormer/blob/main/video_classification/slowfast/models/uniformer.py就可以了

  #130 opened 6 months ago by BugMaker2002
  1

• Unknown model (uniformer_large_ls)

  #127 opened 8 months ago by ZZ00325
  2

• What are these "tools/submit.py" and "tools/summit.py" for ?

  #129 opened 8 months ago by Dev-ori
  1

• image_classification_model(Large resolution fine-tuning (384x384))

  #128 opened 8 months ago by wanzhen666
  3

• some ambiguousness when testing on k400

  #126 opened 9 months ago by LMMMEng
  1

• video dataset

  #125 opened 9 months ago by LMMMEng
  2

• which pretrain model is uniformer-small-dim64？

  #124 opened 9 months ago by wanzhen666
  5

• The picture has to be pre-trained

  #123 opened 9 months ago by Aaron198513
  1

• Question regarding Imagenet pretraining

  #122 opened a year ago by MLDeS
  4

• mmdet object detection pretrained model

  #121 opened a year ago by 2252033991
  1

• About the pretrain model

  #114 opened a year ago by BaoliangChen-stu
  3

• Usage of Image classification with UniFormer with Token Labeling

  #119 opened a year ago by shimmer-ghq
  1

• Error in loading videos for testing

  #117 opened a year ago by mr17m
  7

• 代码问题

  #112 opened a year ago by OvO616
  4

• Welcome update to OpenMMLab 2.0

  #109 opened a year ago by vansin
  1

• Thank you very much! I found the answer in the previous answer!

  #120 opened a year ago by xupeijun01
  0

• Is pos_embed NEED for every Block？

  #118 opened a year ago by mywebinfo65536
  2

• SABlock is the same as SABlock_windows？

  #116 opened a year ago by mywebinfo65536
  2

• About the code for ucf101 dataset processing

  #115 opened a year ago by yult0821
  2

• The model and log links for the ucf101 and hmdb51 datasets are no longer valid

  #113 opened a year ago by Chunsheng13
  3

• One 7x7 conv vs. two 3x3 conv

  #111 opened a year ago by LMMMEng
  4

• pos_embed key missing from checkpoints

  #110 opened a year ago by schmitzmatheus
  1

• Overflow risk when calculate softmax?

  #108 opened a year ago by realgump
  2

• Loss&error explosion

  #106 opened 2 years ago by SKBL5694
  1

• spatiotemporal behavior detection

  #107 opened 2 years ago by yan-ctrl
  10

• Found a new paper from you: InternVideo

  #105 opened 2 years ago by SKBL5694
  1

• 训练视频模型时如何正确加载Imagenet的预训练权重

  #102 opened 2 years ago by LEM0NTE
  3

• How to process video data into the form of new_k400?

  #104 opened 2 years ago by SKBL5694
  1

• Failed to meta load video idx XXXX from PATH when running 'bash exp/uniformer_s8x8_k400/run.sh' from the Uniformer/video_classification directory

  #103 opened 2 years ago by SWE-Ahmed
  1

• Hope to get a copy of the data stored in BaiduYun

  #101 opened 2 years ago by SKBL5694
  5

• 请教一些写代码方面的问题

  #100 opened 2 years ago by SKBL5694
  3

• Training with custom dataset

  #91 opened 2 years ago by manhcntt21
  4

• Training time for kinetics-400

  #96 opened 2 years ago by AbdelrahmanShakerYousef
  6

• New dataset

  #97 opened 2 years ago by LEM0NTE
  6

• Training from n epoch

  #98 opened 2 years ago by manhcntt21
  0

• K400上预训练的权重，如何在sthv2微调训练时正确加载

  #94 opened 2 years ago by LEM0NTE
  3

• something-something V1 dataset unavailable

  #95 opened 2 years ago by LEM0NTE
  1

• Test result lower than paper

  #92 opened 2 years ago by LEM0NTE
  4

• About the params setting of use_checkpoint

  #93 opened 2 years ago by go-ahead-maker
  5

• How to draw the figure 8?

  #89 opened 2 years ago by hongsheng-Z
  4

• Structure design for Video

  #90 opened 2 years ago by LEM0NTE
  5

• Test on kinetics 400

  #86 opened 2 years ago by manhcntt21
  6

• a basic question about pretrained model

  #87 opened 2 years ago by go-ahead-maker
  4

• KeyError: 'pos_embed' when fine-tuning

  #88 opened 2 years ago by RooKichenn
  3

• Error when clone

  #84 opened 2 years ago by manhcntt21
  4