SysCV/qdtrack

Issues

• qdtrack based on YOLOX detector.

  #135 opened 2 years ago by laybebe
  4

• The model and loaded state dict do not match exactly during inference

  #137 opened 2 years ago by Faheem-Yousafzai
  2

• Model links expired :(

  #143 opened a year ago by thisistejaspandey
  0

• Where is ckpts/mmdet/faster_rcnn_r50_caffe_fpn_person_ap551.pth?

  #142 opened a year ago by JackWoo0831
  0

• Provided checkpoint does NOT have weights for the Embedding head

  #138 opened 2 years ago by tanjary21
  6

• Test on MOT17, ModuleNotFoundError: No module named 'qdtrack'

  #113 opened 2 years ago by iTruffle
  2

• Using detector only

  #141 opened 2 years ago by Raspberry-beans
  0

• AttributeError: 'list' object has no attribute 'items'

  #107 opened 3 years ago by animesh-007
  2

• Issue in conversion to BDD100K format

  #139 opened 2 years ago by Raspberry-beans
  0

• The model and loaded state dict do not match exactly

  #125 opened 2 years ago by yimingzhou1
  3

• How to change the Faster rcnn detector to Mask R-CNN ?

  #136 opened 2 years ago by lijoe123
  0

• QDTrack based on YOLO-X detector

  #131 opened 2 years ago by sanghoooon
  3

• Welcome update to OpenMMLab 2.0

  #133 opened 2 years ago by vansin
  0

• Question about training settings when submitting to MOT test server

  #122 opened 2 years ago by sanghoooon
  0

• detector freeze problem

  #128 opened 2 years ago by YOOHYOJEONG
  1

• Can I train only tracker?

  #126 opened 2 years ago by YOOHYOJEONG
  2

• Do you support HOTA metric ?

  #123 opened 2 years ago by foolhard
  1

• Confusion on the relationship among samples_per_GPU, nums of GPUs and LR

  #119 opened 2 years ago by jkd2021
  1

• BDD100k det conversion error

  #121 opened 2 years ago by IMcDougall
  0

• Problem while converting BDD tracking dataset. KeyError: "index" and KeyError : "Occluded", KeyError: 'Truncated'

  #120 opened 2 years ago by YOOHYOJEONG
  3

• The reference image and key image are exactly the same

  #118 opened 2 years ago by HuseyinVisea
  4

• Training accuracy meaning

  #96 opened 2 years ago by AmanGoyal99
  1

• from ..evaluation import xyxy2xywh

  #84 opened 2 years ago by taheranjary
  0

• Any ideas to put on JDE / FairMot or other one-hot trackers?

  #94 opened 2 years ago by long2double
  0

• gradients exploding

  #87 opened 2 years ago by SparksDing
  1

• TypeError : Resnet : __init__() got an unexpected keyword argument 'init_cfg'

  #90 opened 2 years ago by AmanGoyal99
  12

• Possibility of using pre-trained weights for neck and head and only customizing backbone ?

  #93 opened 2 years ago by AmanGoyal99
  1

• Backbone pre-trained on BDD100K dataset

  #95 opened 2 years ago by AmanGoyal99
  2

• Possibility of using pretrained weights for neck and heads ?

  #100 opened 2 years ago by AmanGoyal99
  0

• Waymo config release?

  #102 opened 2 years ago by carlhuang070
  0

• Waymo MOT support

  #101 opened 2 years ago by carlhuang070
  0

• unexpected key in source state_dict: rpn_head.rpn_conv.weight, rpn_head.rpn_conv.bias, rpn_head.rpn_cls.weight, rpn_head.rpn_cls.bias, rpn_head.rpn_reg.weight, rpn_head.rpn_reg.bias

  #103 opened 2 years ago by AmanGoyal99
  0

• Inference

  #105 opened 2 years ago by LisaBernhardt
  1

• Example code for inference?

  #104 opened 2 years ago by sarimmehdi
  1

• How do I get the det_${SET_NAME}. json file?

  #109 opened 2 years ago by QiyuLuo
  1

• Whether QDTrack supports changing detectors or using external detection results for tracking

  #116 opened 2 years ago by unihornWwan
  1

• I'm confusing with the meaning of auxiliary loss

  #111 opened 2 years ago by hcv1027
  4

• About backdrops

  #110 opened 3 years ago by Leo63963
  0

• problems when preparing bdd100k dataset

  #106 opened 3 years ago by Zhangyongtao123
  1

• How to do batch inference?

  #88 opened 3 years ago
  3

• Computing Overall Loss

  #79 opened 3 years ago by LisaBernhardt
  1

• Inconsistent Results on BDD100K Tracking Validation Set

  #82 opened 3 years ago by taheranjary
  8

• MOT17 Exploding Training Gradients

  #81 opened 3 years ago by taheranjary
  3

• 'QDTrack' object has no attribute 'tracker'

  #97 opened 3 years ago by gurkirt
  2

• Is customization of backbone possible as mentioned in the mmdet library ?

  #91 opened 3 years ago by AmanGoyal99
  5

• KeyError: 'frame_id' when using inference_model from inference api

  #92 opened 3 years ago
  1

• Where is tao-classes.txt?

  #89 opened 3 years ago
  1

• Is it possible to drop the unmatched embedding features?

  #80 opened 3 years ago
  3

• Your tracker doesn't perform well in the wild, please add the ability to drop embeddings

  #85 opened 3 years ago
  2

• [Error]: Missing Image during train

  #83 opened 3 years ago by damon-demon
  1