open-mmlab/mmaction2

Issues

• [Bug] different performance about batch size on val pipeline and test pipeline

  #2801 opened 3 months ago by rkdckddnjs9
  1

• [Bug] use tools/misc/clip_feature_extraction.py to extract feature from long video, dataloaders do not work

  #2850 opened 8 days ago by fucker255
  1

• [Docs] Custom Models?

  #2798 opened 3 months ago by SenmiaoORZ
  1

• [Bug] demo.py does not work with MMCV2.2.0

  #2852 opened 6 days ago by rmuraix
  1

• Can adjust batch size on demo skeleton?

  #2851 opened 8 days ago by spped2000
  0

• [Bug] In project "gesture recognition",the number of keypoints output by MMPOSE is inconsistent with the number of keypoints input by MMACTION

  #2849 opened 9 days ago by wangbofei11
  0

• CUDA error: CUBLAS_STATUS_EXECUTION_FAILED when calling `cublasSgemm

  #2793 opened 3 months ago by learn-learn202
  1

• [Bug] long-video-demo raised AssertionError on assert crop_h == img_h or crop_w == img_w

  #2783 opened 4 months ago by vba34520
  8

• [Bug]

  #2848 opened 18 days ago by Sunshulong
  0

• [Feature] How to get flops of the MMRecognizer3D?

  #2794 opened 3 months ago by wukurua
  1

• [Bug] Which demo supports reasoning rgbposec3d?

  #2847 opened 25 days ago by aonihao1
  0

• [Bug] ModuleNotFoundError: No module named 'mmaction.registry'

  #2845 opened a month ago by dqj5182
  0

• [Docs] simplest way to check performance of mmaction2 with pretrained model

  #2842 opened a month ago by sourabCV
  1

• GradCAM utils only support Recognizer2D & Recognizer3D.

  #2840 opened a month ago by Ritatanz
  0

• [Docs] How to fine tune BMN when using ActionNet data prep method 2

  #2788 opened a month ago by valentin-fngr
  3

• Impossibility to understand how to create and train custom dataset[Docs]

  #2839 opened a month ago by gguzzy
  0

• KeyError: 'ActionVisualizer is not in the mmengine::visualizer registry.[Bug]

  #2837 opened a month ago by berengueradrian
  0

• 训练时空动作检测 ，如何处理类别不均衡的问题

  #2797 opened 3 months ago by yuweiApp
  1

• vis_cam.py not work in timesformer

  #2835 opened a month ago by whiteswordLI
  1

• [Bug] Training MVit with RawFrameDataset type

  #2834 opened 2 months ago by maximefuchs
  0

• [Bug] batch_size problem

  #2832 opened 2 months ago by Sunshulong
  0

• [Feature] Export the tsm_mobilnetv2 model to onnx format

  #2829 opened 2 months ago by san-kou7
  0

• top1_acc is allways times 0f 0.5or0.00[Bug]

  #2827 opened 2 months ago by Sunshulong
  0

• train[Bug]

  #2826 opened 2 months ago by Sunshulong
  0

• train time[Bug]

  #2825 opened 2 months ago by Sunshulong
  0

• [Bug] File 'visualize_heatmap_volume' does not exist

  #2796 opened 3 months ago by da13132
  3

• [Bug] VisualizationHook is not set in the `default_hooks` field of config. Please set `visualization=dict(type="VisualizationHook")`

  #2824 opened 2 months ago by Qinziqi1996
  0

• Do we now support the method of Skeleton-based Spatio-Temporal Action Detection and Action Recognition

  #2823 opened 2 months ago by SCsleepy
  2

• [Bug] It is not possible to use "analysis_tools\get_flops.py" to get the number of parameters and calculations

  #2822 opened 2 months ago by aonihao1
  3

• [Bug]

  #2821 opened 2 months ago by maple0leaves
  0

• [Docs]

  #2820 opened 2 months ago by LiYH1234
  0

• [Bug] Fail to use confusion_matrix to RGBPoseC3D

  #2815 opened 2 months ago by HoBeom
  3

• [Bug] resize_videos.py : get invalid literal for int() with base 10: ''

  #2814 opened 2 months ago by automatic114
  1

• [Docs] How to train a model with different inputs from different model feature and videos

  #2805 opened 3 months ago by zeynepgokce
  1

• [Docs]

  #2809 opened 3 months ago by iamhamzahussain
  1

• model zoo lost

  #2804 opened 3 months ago by fengjingchehu
  0

• [Bug] Use mmaction2 - main/configs/skeleton/posec3d/rgbpose_conv3d/rgbpose_conv3d py error for the demo

  #2803 opened 3 months ago by aonihao1
  0

• [Docs] Custom Dataset Training with PoseC3D

  #2800 opened 3 months ago by mohammed-alaa40123
  2

• [Docs]How to add YOLO target detection frame to frame people when reasoning in TimeSformer

  #2799 opened 3 months ago by xqc-qc
  0

• Fast inference option for MMAction2 models, like SlowFast

  #2791 opened 3 months ago by KooroshRH
  1

• [Feature]

  #2790 opened 4 months ago by roykev
  0

• [Docs] Custom action localization dataset. anet_anno_train.json missing

  #2787 opened 4 months ago by valentin-fngr
  0

• [Docs] ACRN training on AVA dataset at 15 fps

  #2785 opened 4 months ago by LorBordin
  0

• Installation fails on macbook m2 [Docs]

  #2784 opened 4 months ago by lizaibeim
  0

• [Docs] 20分钟了解 MMAction2 框架设计，步骤2导入问题

  #2779 opened 5 months ago by everglowgh
  0

• [Docs] Fine-tuning ActionClip

  #2778 opened 5 months ago by CarmelaCalabrese
  0

• [Bug] How can I use clip_feature_extraction.py to input the rawframes , and then output the RGB features of each frame?

  #2777 opened 5 months ago by Chenhongchang
  0

• 找不到rtmpose的ckpt[Bug]

  #2775 opened 5 months ago by shiyuxin1999S
  0

• [Feature] Rename the confusing argument 'format_shape' in ActionDataPreprocessor

  #2773 opened 5 months ago by makecent
  0

• get_flops.py[Bug]

  #2772 opened 6 months ago by ChenZhenGui
  0