[AAAI 2024] DiffusionTrack: Diffusion Model For Multi-Object Tracking. DiffusionTrack is the first work to employ the diffusion model for multi-object tracking by formulating it as a generative noise-to-tracking diffusion process.
License: Other
Hi, author. I am very interested in your work, but I have encountered some issues while reading the code below:
def diffusion_postprocess(self,diffusion_outputs,conf_scores,nms_thre=0.7,conf_thre=0.6):
# input shape: (2B, num_proposal, 5), (B, num_proposal, 1)
pre_prediction,cur_prediction=diffusion_outputs.split(len(diffusion_outputs)//2,dim=0)
# (B, num_proposal, 5), (B, num_proposal, 5)
output = [None for _ in range(len(pre_prediction))] # [None, None, ... ...] shape=(B)
for i,(pre_image_pred,cur_image_pred,association_score) in enumerate(zip(pre_prediction,cur_prediction,conf_scores)):
association_score=association_score.flatten() # (num_proposal)
# If none are remaining => process next image
if not pre_image_pred.size(0):
continue
# _, conf_mask = torch.topk((image_pred[:, 4] * class_conf.squeeze()), 1000)
# Detections ordered as (x1, y1, x2, y2, obj_conf, class_conf, class_pred)
detections=torch.zeros((2,len(cur_image_pred),7),dtype=cur_image_pred.dtype,device=cur_image_pred.device)
# detections.shape = (2, num_proposals, 7)
# detections[0]
detections[0,:,:4]=pre_image_pred[:,:4]
detections[1,:,:4]=cur_image_pred[:,:4]
detections[0,:,4]=association_score
detections[1,:,4]=association_score
detections[0,:,5]=torch.sqrt(torch.sigmoid(pre_image_pred[:,4])*association_score)
detections[1,:,5]=torch.sqrt(torch.sigmoid(cur_image_pred[:,4])*association_score)
score_out_index=association_score>conf_thre
# strategy=torch.mean
# value=strategy(detections[:,:,5],dim=0,keepdim=False)
# score_out_index=value>conf_thre
detections=detections[:,score_out_index,:]
if not detections.size(1):
output[i]=detections
continue
nms_out_index_3d = cluster_nms(
detections[0,:,:4],
detections[1,:,:4],
# value[score_out_index],
detections[0,:,4],
iou_threshold=nms_thre)
detections = detections[:,nms_out_index_3d,:]
if output[i] is None:
output[i] = detections
else:
output[i] = torch.cat((output[i], detections))
# output[i] = (2, num_proposal, 7)
return output[0][0],output[0][1],torch.cat([output[1][0],output[1][1]],dim=0) if len(output)>=2 else None
For the diffusion_postprocess function, given that it takes diffusion_outputs (2*batch_size, num_proposal, 5) and conf_scores (batch_size, num_proposal, 1)as inputs, the return value output[0][0] (num_proposal, 7)๏ผoutput[0][1] (num_proposal, 7) ... ... represents the detection results of the previous image for the first/second iteration in the batch. This seems to discard the majority of the information in the batch. Why is the process handled this way? If there's a misunderstanding on my part, please correct me. Much appreciated, thank you!
Hello author, I am confused about the following code. Can you please explain the meaning of this?
2023-11-01 20:36:21.464 | INFO | yolox.core.trainer:before_train:129 - args: Namespace(batch_size=2, ckpt='../pretrained/bytetrack_ablation.pth.tar', devices=1, dist_backend='nccl', dist_url=None, exp_file='../exps/example/mot/yolox_x_diffusion_det_mot17_ablation.py', experiment_name='yolox_x_diffusion_det_mot17_ablation', fp16=False, local_rank=0, machine_rank=0, name=None, num_machines=1, occupy=True, opts=[], resume=False, start_epoch=None)
2023-11-01 20:36:21.465 | INFO | yolox.core.trainer:before_train:130 - exp value:
โโโโโโโโโโโโโโโโโโโโคโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ keys โ values โ
โโโโโโโโโโโโโโโโโโโโชโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโก
โ seed โ 8823 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ output_dir โ './DiffusionTrack_outputs' โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ print_interval โ 20 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ eval_interval โ 5 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ num_classes โ 1 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ depth โ 1.33 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ width โ 1.25 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ data_num_workers โ 4 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ input_size โ (800, 1440) โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ random_size โ (18, 32) โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ train_ann โ 'train_half.json' โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ val_ann โ 'val_half.json' โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ degrees โ 10.0 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ translate โ 0.1 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ scale โ (0.1, 2) โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ mscale โ (0.8, 1.6) โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ shear โ 2.0 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ perspective โ 0.0 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ enable_mixup โ True โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ warmup_epochs โ 1 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ max_epoch โ 30 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ warmup_lr โ 0 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ basic_lr_per_img โ 1.5625e-05 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ scheduler โ 'yoloxwarmcos' โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ no_aug_epochs โ 10 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ min_lr_ratio โ 0.05 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ ema โ True โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ weight_decay โ 0.0005 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ momentum โ 0.9 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ exp_name โ 'yolox_x_diffusion_det_mot17_ablation' โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ test_size โ (800, 1440) โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ random_flip โ False โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ task โ 'detection' โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ conf_thresh โ 0.4 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ det_thresh โ 0.7 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ nms_thresh2d โ 0.75 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ nms_thresh3d โ 0.7 โ
โโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ interval โ 5 โ
โโโโโโโโโโโโโโโโโโโโงโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
2023-11-01 20:36:24.553 | INFO | yolox.data.datasets.mot:init:39 - loading annotations into memory...
2023-11-01 20:36:24.640 | INFO | yolox.data.datasets.mot:init:39 - Done (t=0.09s)
2023-11-01 20:36:24.641 | INFO | pycocotools.coco:init:86 - creating index...
2023-11-01 20:36:24.649 | INFO | pycocotools.coco:init:86 - index created!
2023-11-01 20:36:24.773 | INFO | yolox.core.trainer:before_train:153 - init prefetcher, this might take one minute or less...
2023-11-01 20:36:30.630 | INFO | yolox.data.datasets.mot:init:39 - loading annotations into memory...
2023-11-01 20:36:30.735 | INFO | yolox.data.datasets.mot:init:39 - Done (t=0.10s)
2023-11-01 20:36:30.735 | INFO | pycocotools.coco:init:86 - creating index...
2023-11-01 20:36:30.744 | INFO | pycocotools.coco:init:86 - index created!
2023-11-01 20:36:30.857 | INFO | yolox.core.trainer:before_train:181 - Training start...
2023-11-01 20:36:30.857 | INFO | yolox.core.trainer:before_epoch:192 - ---> start train epoch1
2023-11-01 20:36:31.669 | INFO | yolox.core.trainer:after_train:187 - Training of experiment is done and the best AP is 0.00
2023-11-01 20:36:31.670 | ERROR | yolox.core.launch:launch:90 - An error has been caught in function 'launch', process 'MainProcess' (38847), thread 'MainThread' (140253388433216):
Traceback (most recent call last):
File "/home/wangtuo/Downloads/Multi-Object-Tracking/Paper/DiffusionTrack-main/tools/train.py", line 133, in
args=(exp, args),
โ โ Namespace(batch_size=2, ckpt='../pretrained/bytetrack_ablation.pth.tar', devices=1, dist_backend='nccl', dist_url=None, exp_f...
โ โโโโโโโโโโโโโโโโโโโโคโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ...
File "/home/wangtuo/Downloads/Multi-Object-Tracking/Paper/DiffusionTrack-main/yolox/core/launch.py", line 90, in launch
main_func(*args)
โ โ (โโโโโโโโโโโโโโโโโโโโคโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ...
โ <function main at 0x7f8f34e2ec80>
File "/home/wangtuo/Downloads/Multi-Object-Tracking/Paper/DiffusionTrack-main/tools/train.py", line 110, in main
trainer.train()
โ โ <function Trainer.train at 0x7f8f44900840>
โ <yolox.core.trainer.Trainer object at 0x7f8f34e38a20>
File "/home/wangtuo/Downloads/Multi-Object-Tracking/Paper/DiffusionTrack-main/yolox/core/trainer.py", line 73, in train
self.train_in_epoch()
โ โ <function Trainer.train_in_epoch at 0x7f8f44905620>
โ <yolox.core.trainer.Trainer object at 0x7f8f34e38a20>
File "/home/wangtuo/Downloads/Multi-Object-Tracking/Paper/DiffusionTrack-main/yolox/core/trainer.py", line 82, in train_in_epoch
self.train_in_iter()
โ โ <function Trainer.train_in_iter at 0x7f8f4491ebf8>
โ <yolox.core.trainer.Trainer object at 0x7f8f34e38a20>
File "/home/wangtuo/Downloads/Multi-Object-Tracking/Paper/DiffusionTrack-main/yolox/core/trainer.py", line 88, in train_in_iter
self.train_one_iter()
โ โ <function Trainer.train_one_iter at 0x7f8f4491ec80>
โ <yolox.core.trainer.Trainer object at 0x7f8f34e38a20>
File "/home/wangtuo/Downloads/Multi-Object-Tracking/Paper/DiffusionTrack-main/yolox/core/trainer.py", line 105, in train_one_iter
outputs = self.model(inps,targets,self.random_flip,self.input_size)
โ โ โ โ โ โ โ โ (800, 1440)
โ โ โ โ โ โ โ <yolox.core.trainer.Trainer object at 0x7f8f34e38a20>
โ โ โ โ โ โ False
โ โ โ โ โ <yolox.core.trainer.Trainer object at 0x7f8f34e38a20>
โ โ โ โ (tensor([[[ 0.0000, 1009.0435, 306.2504, 15.8012, 31.7075],
โ โ โ [ 0.0000, 992.1570, 280.6044, 14.5298, 27...
โ โ โ (tensor([[[[-0.8335, -0.8335, -0.8335, ..., 1.2282, 1.1493, 1.1380],
โ โ [-0.8335, -0.8335, -0.8335, ..., 1.2282,...
โ โ DiffusionNet(
โ (backbone): YOLOPAFPN(
โ (backbone): CSPDarknet(
โ (stem): Focus(
โ (conv): BaseConv(
โ (...
โ <yolox.core.trainer.Trainer object at 0x7f8f34e38a20>
File "/home/wangtuo/anaconda3/envs/diffusionTrack/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
return forward_call(*input, **kwargs)
โ โ โ {}
โ โ ((tensor([[[[-0.8335, -0.8335, -0.8335, ..., 1.2282, 1.1493, 1.1380],
โ [-0.8335, -0.8335, -0.8335, ..., 1.2282...
โ <bound method DiffusionNet.forward of DiffusionNet(
(backbone): YOLOPAFPN(
(backbone): CSPDarknet(
(stem): Focus(...
File "/home/wangtuo/Downloads/Multi-Object-Tracking/Paper/DiffusionTrack-main/diffusion/models/diffusionnet.py", line 77, in forward
features,mate_info,targets=torch.cat([pre_targets,cur_targets],dim=0))
โ โ โ โ โ โ tensor([[[ 0.0000, 1009.0435, 306.2504, 15.8012, 31.7075],
โ โ โ โ โ [ 0.0000, 992.1570, 280.6044, 14.5298, 27....
โ โ โ โ โ tensor([[[ 0.0000, 1009.0435, 306.2504, 15.8012, 31.7075],
โ โ โ โ [ 0.0000, 992.1570, 280.6044, 14.5298, 27....
โ โ โ โ <built-in method cat of type object at 0x7f8f30351e80>
โ โ โ <module 'torch' from '/home/wangtuo/anaconda3/envs/diffusionTrack/lib/python3.7/site-packages/torch/init.py'>
โ โ (torch.Size([2, 3, 800, 1440]), device(type='cuda', index=0), torch.float32)
โ ([tensor([[[[-1.9567e-02, -9.6989e-02, -2.6508e-01, ..., -2.2024e-01,
-2.7844e-01, -2.7244e-01],
[ 4.84...
File "/home/wangtuo/anaconda3/envs/diffusionTrack/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
return forward_call(*input, **kwargs)
โ โ โ {'targets': tensor([[[ 0.0000, 1009.0435, 306.2504, 15.8012, 31.7075],
โ โ [ 0.0000, 992.1570, 280.6044, 14...
โ โ (([tensor([[[[-1.9567e-02, -9.6989e-02, -2.6508e-01, ..., -2.2024e-01,
โ -2.7844e-01, -2.7244e-01],
โ [ 4.8...
โ <bound method DiffusionHead.forward of DiffusionHead(
(head): DynamicHead(
(box_pooler): ROIPooler(
(level_pooler...
File "/home/wangtuo/Downloads/Multi-Object-Tracking/Paper/DiffusionTrack-main/diffusion/models/diffusion_head.py", line 388, in forward
loss_dict = self.criterion(output, targets)
โ โ โ [{'labels': tensor([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
โ โ 0, 0, 0, 0, 0, 0, 0, 0, 0...
โ โ {'pred_logits': tensor([[[-2.6417],
โ [-3.3217],
โ [-3.2851],
โ ...,
โ [-3.0609],
โ [-2.90...
โ DiffusionHead(
(head): DynamicHead(
(box_pooler): ROIPooler(
(level_poolers): ModuleList(
(0): ROIAlign(o...
File "/home/wangtuo/anaconda3/envs/diffusionTrack/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
return forward_call(*input, **kwargs)
โ โ โ {}
โ โ ({'pred_logits': tensor([[[-2.6417],
โ [-3.3217],
โ [-3.2851],
โ ...,
โ [-3.0609],
โ [-2.9...
โ <bound method SetCriterionDynamicK.forward of SetCriterionDynamicK(
(matcher): HungarianMatcherDynamicK()
)>
File "/home/wangtuo/Downloads/Multi-Object-Tracking/Paper/DiffusionTrack-main/diffusion/models/diffusion_losses.py", line 233, in forward
indices, _ = self.matcher(outputs_without_aux, targets)
โ โ โ [{'labels': tensor([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
โ โ 0, 0, 0, 0, 0, 0, 0, 0, 0...
โ โ {'pred_logits': tensor([[[-2.6417],
โ [-3.3217],
โ [-3.2851],
โ ...,
โ [-3.0609],
โ [-2.90...
โ SetCriterionDynamicK(
(matcher): HungarianMatcherDynamicK()
)
File "/home/wangtuo/anaconda3/envs/diffusionTrack/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
return forward_call(*input, **kwargs)
โ โ โ {}
โ โ ({'pred_logits': tensor([[[-2.6417],
โ [-3.3217],
โ [-3.2851],
โ ...,
โ [-3.0609],
โ [-2.9...
โ <bound method HungarianMatcherDynamicK.forward of HungarianMatcherDynamicK()>
File "/home/wangtuo/Downloads/Multi-Object-Tracking/Paper/DiffusionTrack-main/diffusion/models/diffusion_losses.py", line 383, in forward
cost_giou = -generalized_box_iou(bz_boxes_pre,bz_boxes_curr,bz_gtboxs_abs_xyxy_pre,bz_gtboxs_abs_xyxy_curr)
โ โ โ โ โ tensor([[1001.1429, 290.3966, 1016.9442, 322.1042],
โ โ โ โ [ 984.8921, 266.8271, 999.4219, 294.3816],
โ โ โ โ [1068.397...
โ โ โ โ tensor([[1001.1429, 290.3966, 1016.9442, 322.1042],
โ โ โ [ 984.8921, 266.8271, 999.4219, 294.3816],
โ โ โ [1068.397...
โ โ โ tensor([[ 480.7337, -1106.8778, 480.7337, -1106.8778],
โ โ [ 817.8951, -325.3340, 817.8951, -325.3340],
โ โ [...
โ โ tensor([[ 905.1316, 297.8546, 905.1316, 297.8546],
โ [ 964.9771, 156.5497, 964.9771, 156.5497],
โ [ 727.037...
โ <function generalized_box_iou at 0x7f8f32cb07b8>
File "/home/wangtuo/Downloads/Multi-Object-Tracking/Paper/DiffusionTrack-main/yolox/utils/box_ops.py", line 77, in generalized_box_iou
return giou_3d(boxes1,boxes3,boxes2,boxes4)
โ โ โ โ โ tensor([[1001.1429, 290.3966, 1016.9442, 322.1042],
โ โ โ โ [ 984.8921, 266.8271, 999.4219, 294.3816],
โ โ โ โ [1068.397...
โ โ โ โ tensor([[ 480.7337, -1106.8778, 480.7337, -1106.8778],
โ โ โ [ 817.8951, -325.3340, 817.8951, -325.3340],
โ โ โ [...
โ โ โ tensor([[1001.1429, 290.3966, 1016.9442, 322.1042],
โ โ [ 984.8921, 266.8271, 999.4219, 294.3816],
โ โ [1068.397...
โ โ tensor([[ 905.1316, 297.8546, 905.1316, 297.8546],
โ [ 964.9771, 156.5497, 964.9771, 156.5497],
โ [ 727.037...
โ <function giou_3d at 0x7f8f40e481e0>
File "/home/wangtuo/Downloads/Multi-Object-Tracking/Paper/DiffusionTrack-main/yolox/utils/cluster_nms.py", line 70, in giou_3d
intercd = intersect(box_c,box_d)
โ โ โ tensor([[[1001.1429, 290.3966, 1016.9442, 322.1042],
โ โ [ 984.8921, 266.8271, 999.4219, 294.3816],
โ โ [1068....
โ โ tensor([[[ 480.7337, -1106.8778, 480.7337, -1106.8778],
โ [ 817.8951, -325.3340, 817.8951, -325.3340],
โ ...
โ <torch.jit.ScriptFunction object at 0x7f8f40e4c258>
RuntimeError: nvrtc: error: invalid value for --gpu-architecture (-arch)
nvrtc compilation failed:
#define NAN __int_as_float(0x7fffffff)
#define POS_INFINITY __int_as_float(0x7f800000)
#define NEG_INFINITY __int_as_float(0xff800000)
template
device T maximum(T a, T b) {
return isnan(a) ? a : (a > b ? a : b);
}
template
device T minimum(T a, T b) {
return isnan(a) ? a : (a < b ? a : b);
}
extern "C" global
void fused_min_max_sub_clamp(float* t_, float* t__, float* t___, float* t____, float* aten_clamp) {
{
if (512 * blockIdx.x + threadIdx.x<61000 ? 1 : 0) {
float t____1 = ldg(t_ + (512 * blockIdx.x + threadIdx.x) % 2 + 4 * (((512 * blockIdx.x + threadIdx.x) / 122) % 500));
float t_____1 = ldg(t__ + (512 * blockIdx.x + threadIdx.x) % 2 + 4 * (((512 * blockIdx.x + threadIdx.x) / 2) % 61));
float t__1 = _ldg(t + (512 * blockIdx.x + threadIdx.x) % 2 + 4 * (((512 * blockIdx.x + threadIdx.x) / 122) % 500));
float t___1 = ldg(t + (512 * blockIdx.x + threadIdx.x) % 2 + 4 * (((512 * blockIdx.x + threadIdx.x) / 2) % 61));
aten_clamp[512 * blockIdx.x + threadIdx.x] = (minimum(t_____1,t____1)) - (maximum(t___1,t__1))<0.f ? 0.f : (minimum(t_____1,t____1)) - (maximum(t___1,t__1));
}
}
}
Dear author thank you very much for your excellent work! I can not find "diffusiontrack_ablation_det.pth.tar".Can you help me?Thank you!
ๅฆ้ข๏ผ่ฟๆๅคงไฝฌๆๆ
Hello author, your work is excellent๏ผ
I couldn't find the parts about using the Box Padding Strategy and Perturbation Schedule while looking through the code.
Could you please let me know in which file I can find these?
When I read the code, I found that due to โself.sampling_steps=1โ during tracking, the model will not enter ddimโs โdef diffusion(sampling_times,bboxes,x_start,pred_noise)โ. What is the authorโs intention of doing this?
Looking forward to your reply
ERROR | yolox.core.launch:90 - An error has been caught in function 'launch', process 'MainProcess' (44024), thread 'MainThread' (540):
ๆจๅฅฝ๏ผๆๅจDancetrackไธ่ฟ่ก่ฎญ็ปๆต่ฏ๏ผ็ฒพๅบฆๅฅไฝๆ ๆฏ๏ผHOTAๅชๆ23.7๏ผๅๆถๆๅฉ็จๆจๆไพ็checkpoint่ฟ่กinference็ฒพๅบฆไนๅชๆ46.3
Dear author thank you very much for your excellent work! I can not find "tools/mix_data_test_dancetrack.py".Can you help me?Thank you!
Dear author thank you very much for your excellent work! Are files such as /datasets missing in yolox/data/ ? I encountered missing 'DiffusionMosaicDetection' from 'yolox.data' and 'MosaicDetection' from 'yolox.data' while training
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
Django
The Web framework for perfectionists with deadlines.
Laravel
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft
Open source projects and samples from Microsoft.
Google
Google โค๏ธ Open Source for everyone.
Alibaba
Alibaba Open Source for everyone
D3
Data-Driven Documents codes.
Tencent
China tencent open source team.