chen-si-jia / Trajectory-Long-tail-Distribution-for-MOT

⭕️ Official codes for "Delving into the Trajectory Long-tail Distribution for Muti-object Tracking" (CVPR2024)
MIT License
41 stars 6 forks source link

Cannot find SVA and DVA methods. #1

Open hxdflying opened 3 months ago

hxdflying commented 3 months ago

Thank you so much for your wonderful work. Where are the SVA and DVA methods in the code, please?

chen-si-jia commented 3 months ago

The SVA and DVA code will be updated soon. They are not difficult to implement. If you are in a hurry, you can implement them according to the description in the paper.

wenkebei commented 2 months ago

Thank you so much for your wonderful work.where are Input:id_target_all_nID_MOT17_half_train.txt mean in GS/step1,please?

chen-si-jia commented 2 months ago

Thank you so much for your wonderful work.where are Input:id_target_all_nID_MOT17_half_train.txt mean in GS/step1,please?

id_target_all_nID_MOT17_half_train.txt is the id label for the training data. It can be obtained by uncommenting lines 53-54, 85-86 and 120-123 of the src/lib/trains/base_trainer.py file.

chen-si-jia commented 2 months ago

Thank you so much for your wonderful work.where are Input:id_target_all_nID_MOT17_half_train.txt mean in GS/step1,please?

id_target_all_nID_MOT17_half_train.txt is the id label for the training data. It can be obtained by uncommenting lines 53-54, 85-86 and 120-123 of the src/lib/trains/base_trainer.py file.

Lsyccnu commented 3 weeks ago

Thank you so much for your wonderful work.where are Input:id_target_all_nID_MOT17_half_train.txt mean in GS/step1,please?

id_target_all_nID_MOT17_half_train.txt is the id label for the training data. It can be obtained by uncommenting lines 53-54, 85-86 and 120-123 of the src/lib/trains/base_trainer.py file. 您好,请问这个获取是需要怎么运行base_trainer.py 文件呢?这貌似是一个封装训练方法的类?

chen-si-jia commented 3 weeks ago

Thank you so much for your wonderful work.where are Input:id_target_all_nID_MOT17_half_train.txt mean in GS/step1,please?

id_target_all_nID_MOT17_half_train.txt is the id label for the training data. It can be obtained by uncommenting lines 53-54, 85-86 and 120-123 of the src/lib/trains/base_trainer.py file. 您好,请问这个获取是需要怎么运行base_trainer.py 文件呢?这貌似是一个封装训练方法的类?

base_trainer.py 不是主函数程序,不可以单独运行。需要运行训练代码(训练指令位于experiments文件夹下),就可以运行进去。

wwwshengbo commented 2 weeks ago

SVA and DVA methods @chen-si-jia 你好,论文提出的方法感觉非常好,想问一下这两个方法 SVA and DVA methods,大概什么时候会上传? 没上传,会影响 MOT16_add_our_method_dla34.sh 正常使用?

chen-si-jia commented 2 weeks ago

SVA and DVA methods @chen-si-jia 你好,论文提出的方法感觉非常好,想问一下这两个方法 SVA and DVA methods,大概什么时候会上传? 没上传,会影响 MOT16_add_our_method_dla34.sh 正常使用?

DVA代码刚已开源,位置位于 tackle_module/DVA 文件夹中。MOT16_add_our_method_dla34.sh 完整的话,需要都运用SVA、DVA和GS。SVA代码也简单,是可以从开源的DVA,加一个卡尔曼滤波,拼凑后即可。

chen-si-jia commented 2 weeks ago

DVA代码刚已开源。SVA代码也简单,是可以从开源的DVA,加一个卡尔曼滤波,拼凑后即可。

chen-si-jia commented 2 weeks ago

The DVA code has just opened. The SVA code is also simple, can be from the opened DVA code, adding a Kalman filter. That's it.

wwwshengbo commented 2 weeks ago

DVA代码刚已开源。SVA代码也简单,是可以从开源的DVA,加一个卡尔曼滤波,拼凑后即可。

小白入门刚跟踪,感谢感谢非常感谢

chen-si-jia commented 2 weeks ago

DVA代码刚已开源。SVA代码也简单,是可以从开源的DVA,加一个卡尔曼滤波,拼凑后即可。

小白入门刚跟踪,感谢感谢非常感谢

我们论文最重要的意义是开启一个新问题,可以尝试对我们提出的这个问题进行新的探索,祝你取得一定的成绩。

wwwshengbo commented 2 weeks ago

DVA代码刚已开源。SVA代码也简单,是可以从开源的DVA,加一个卡尔曼滤波,拼凑后即可。

小白入门刚跟踪,感谢感谢非常感谢

我们论文最重要的意义是开启一个新问题,可以尝试对我们提出的这个问题进行新的探索,祝你取得一定的成绩。

确实,感觉论文角度非常好,但对于跨专业,刚入门人来说,代码可能正常运行,都是个问题