- Видео 48
- Просмотров 7 512
Jumabek Alikhanov
Южная Корея
Добавлен 4 июн 2012
Early Stage Intelligent Automatic Parking Management
AI programm enables parking facilities to subscribe to our mobile app and seamlessly start charging for parking use in an intelligent manner
Просмотров: 17
Видео
Another sample video where tracker fails to predict tracks
Просмотров 26Год назад
: VIRAT S 010111 00 000000 000032 visualization Since object size is too small, tracker fails to predict
Issues for using VIRAT videos for MOT task
Просмотров 28Год назад
1. Ground truth BoundingBoxes (green) are losely annotated. 2. video is not fully annotated. In the example, annotation starts after 301s frame in the 813 frame video. This is problematic because, tracker that makes prediction for objects before 301 frames will be penalized unjustly
Importance of adjusting tracker Hyper Prams to environment Part 2/2
Просмотров 64Год назад
Only difference between this and previous video is conf threshold is increased from 0.35 - 0.5
DeepOCSORT: MOT tracking in real-life scenarios seems far from being useful? Part 1/2
Просмотров 195Год назад
DeepOCSORT demo using yolov8_tracking Commands & arguments: python examples/track.py source assets/tests_reid_after_object_disappears_from_frame.avi tracking-method deepocsort yolo-model yolov8m conf 0.35 show save classes 0 Reference github.com/mikel-brostrom/yolo_tracking
ByteTrack on virat. Failure cases : VIRAT S 010208 08 000807 000831
Просмотров 22Год назад
ByteTrack on virat. Failure case This video shows how bytetrack performance is poor on CCTV scenario
ByteTrack on VIRAT CCTV data | Multi Object Tracking | MOT
Просмотров 131Год назад
ByteTrack on VIRAT CCTV data | Multi Object Tracking | MOT
VIRAT, ByteTrack, Multi Object tracking, ground truth, object track visuaalization
Просмотров 69Год назад
VIRAT, ByteTrack, Multi Object tracking, ground truth, object track visuaalization
The result is truly excellent! I tried the YOLOv8 with DeepSort, StrongSort, and ByteTrack algorithms. However, every time the camera detects an object, the IDs change. But Your IDs remain constant. Can you help me with how you achieved this? Can you share a video or a resource? Thank you.
I have a issue similar what you mentioned here, hope you post new video to resolve this problem. How ignore what not moving
We have explained the procedure here ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=10173520 you can also find the cleaned subset of VIRAT - which we name as SurvTrack here github.com/Jumabek/SurvTrack
@@jumabekalikhanov5237 thank you so much
That is cool!
so fast, that is cool)
Good job professor 👏
Source code counting?
👏👏👏👏👏
Hii, Can you make a video of how you did it? or can you share the code? Pleasee
Hi, you need to train pedestrian/people detector, which you can train using pedestrian datasets. This repo shows how github.com/AlexeyAB/darknet
👍👍👍
Hey Jumabek, Good work! Currently, I'm working on a similar project like this. I have some confusion regarding your code. I'm following your method. Is it possible that we can talk about this on skype or WhatsApp? Please contact me. I really need your help. Here is my Gmail account. aisarali99@gmail.com. please Mail me your contact details. Thanks
Men towken kimyo hetnologiyada oqiman magistrni boshqa yonalishda oqisam boladimi?
bo'ladi, agar domla qabul qilsa
Assalomu aleykum aka men instda 4 kursman bu yil bitiraman lekin ilets yoq koreys tilini oqib ham yutsa boladimi javob uchun oldindan raxmat
siz topshirayotgan universitetga bog'liq, ularda kores tili darajasi (TOPIK) uchun grant borligini vebsayt orqali tekshirish kk
Jumabek, gap yo'q do'stim :)
Кайси шахарида укийсиз.кайси университет
Assalomu Aleykum oka man hozr uzbda uqiman 3kursman baholarim 5ga yopiq kelyabman endi IELTS 5.5 Olsam koreya universitetlariga topshirishim mumknmi grantga?
mumkin, lekin IELTS ni 5.5 dan balandroq olsangiz osonroq grant topasiz
Thanks for posting this. Was looking for it then I saw it wasn't on utube anymore.
Sir is this real time, can you help me to make a program like this. please and thanks