ScaledYOLOv4
deep_sort
ScaledYOLOv4 | deep_sort | |
---|---|---|
10 | 10 | |
2,017 | 5,068 | |
- | - | |
0.0 | 0.0 | |
10 months ago | 19 days ago | |
Python | Python | |
GNU General Public License v3.0 only | GNU General Public License v3.0 only |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
ScaledYOLOv4
-
DeepSort with PyTorch(support yolo series)
WongKinYiu/ScaledYOLOv4
-
[P] Oddly thresholded confidence scores on scaled yolov4 csp
I'm using a branch of the author's PyTorch repo
-
Real time object detection and recognition
Take a look at yolov5 or scaled yolov4. They should both handle real-time training, at low enough resolution anyway; I don't know if there is any model that can do real-time detection on 4K videos. Don't pay attention to the version numbers, I think the scaled yolov4 is sliiiightly better performance.
-
YOLOR (Scaled-YOLOv4-based): The best speed/accuracy ratio for Waymo autonomous driving challenge
[CVPR'21 WAD] Challenge - Waymo Open Dataset: https://waymo.com/open/challenges/2021/real-time-2d-prediction/ YOLOR (Scaled-YOLOv4-based) has the best speed/accuracy ratio on Waymo autonomous driving challenge ((Waymo Open Dataset): Real-time 2D Detection. Thanks Chien-Yao Wang from Academia Sinica and DiDi MapVision team to push Scaled-YOLOv4 further! * DIDI MapVision: https://arxiv.org/abs/2106.08713 * YOLOR https://arxiv.org/abs/2105.04206 * YOLOR-code (Pytorch): https://github.com/WongKinYiu/yolor * Scaled-YOLOv4(CVPR21): https://openaccess.thecvf.com/content/CVPR2021/html/Wang\_Scaled-YOLOv4\_Scaling\_Cross\_Stage\_Partial\_Network\_CVPR\_2021\_paper.html * Scaled-YOLOv4-code (Pytorch): https://github.com/WongKinYiu/ScaledYOLOv4 * YOLOv4: https://arxiv.org/abs/2004.10934 * YOLOv4-code (Darknet, Pytorch, TensorFlow, TRT, OpenCV…): https://github.com/AlexeyAB/darknet#yolo-v4-in-other-frameworks
The DiDi MapVision team has shown excellent results with the YOLOR and DIDI MapVision models, both based on Scaled-YOLOv4: * DIDI MapVision: https://arxiv.org/abs/2106.08713 * YOLOR https://arxiv.org/abs/2105.04206 * YOLOR-code (Pytorch): https://github.com/WongKinYiu/yolor * Scaled-YOLOv4(CVPR21): https://openaccess.thecvf.com/content/CVPR2021/html/Wang_Scaled-YOLOv4_Scaling_Cross_Stage_Partial_Network_CVPR_2021_paper.html * Scaled-YOLOv4-code (Pytorch): https://github.com/WongKinYiu/ScaledYOLOv4 * YOLOv4: https://arxiv.org/abs/2004.10934 * YOLOv4-code (Darknet, Pytorch, TensorFlow, TRT, OpenCV...): https://github.com/AlexeyAB/darknet#yolo-v4-in-other-frameworks
-
[P] YOLOR (Scaled-YOLOv4-based): The best speed/accuracy ratio for Waymo autonomous driving challenge
* Scaled-YOLOv4-code (Pytorch): https://github.com/WongKinYiu/ScaledYOLOv4
-
Implementing Faster R-CNN in C
As for the BEST model, I would suggest you use Scaled YOLOv4 since it performs the best both on the cloud and edge devices.
- How do you add a class to coco classes in YOLO (CNN) object detection?
-
How to catch up with trending computer vision open-source Github repos
You're out of sync, yolo-v5 is a controversial naming. And whatever side of the debate you are in state-of-the art is Scaled-Yolo v4: https://github.com/WongKinYiu/ScaledYOLOv4 https://arxiv.org/abs/2011.08036
-
How to Train a Scaled-YOLOv4 Object Detection Model
To complicate matters further, the code for this improvement of YOLOv4 (published by one of the original YOLOv4 authors) is actually a fork of the YOLOv5 repo: https://github.com/WongKinYiu/ScaledYOLOv4/tree/yolov4-csp
deep_sort
-
Similari 0.26.2: MOT framework with Python bindings
Similari is a Rust/Python framework aimed at building sophisticated tracking systems. With Similari, you can develop highly efficient parallelized SORT, DeepSORT, and other sophisticated multiple-object tracking engines.
-
How to integrate DeepSORT with YOLOv8
I'm doing a Python personal project where I'm trying to use YOLOv8 and DeepSORT to detect vehicles from a car's dash cam footage. I succeeded in using YOLOv8 to output the correct bounding boxes by processing each camera frame. However, I tried to add on DeepSORT code, but it made the detection accuracy significantly worse. I'm pretty sure I need to train my own "feature extractor" for DeepSORT to create a new .pb file. I got this information from the deep_sort GitHub link: https://github.com/nwojke/deep_sort. I tried to find resources to do this but they are pretty scarce. Has anyone had experience with this problem?
-
Need to download resources for DeepSORT from pan.baidu.com
The feature model well that looks like it is at that domain you mentioned but why not instead of using this repo use the original authors repo
-
DeepSort with PyTorch(support yolo series)
nwojke/deep_sort
-
Kalman filter in Rust runs 120+ times faster than NumPy, SciKit implementation
I was implementing the Kalman filter for bounding boxes during the last two days. As an inspiration source, I looked at the Python3 Kalman filter implementation that is used in the DeepSORT algorithm and uses NumPy and SciKit under the hood, so it's pretty efficient because all the operations are run inside FFI.
-
[P] The easiest way to process and tag video data
There's tons of work out there when it comes to object tracking such as DeepSort. We've worked to build simpler, more efficient solutions in-house though. Then past that, it's a matter of treating everything in the video as an object (including the whole frame), tracking it, and saving it in a no-SQL DB such that it's easy to query in this way.
-
Building an API + query language for rich data like images and video
Right now, the way we're thinking about it is to turn videos into something that works with the structure of a database like MongoDB. Everything in an image or a video is an object (even the frame itself is an object with a large bounding box), and each of these objects has some attributes and can be tracked over time with some form of object tracking. Given that the objects are tracked, they can each basically be returned as a time-series of each of the attributes associated with that object.
-
Could someone suggest a good article that explains the implementation of deep sort algorithm ?
Deep Sort
-
How do I train the DeepSORT tracker for a custom class?
I was wondering if I could use the same annotated data(in YOLO format) for the training of the tracker as well. I took a look at the original repo for DeepSORT, and it does mention the training using cosine metric learning, but I could not seem to understand how to replicate that for my own dataset(they show us how to do it for the MARS and Market1501 datasets).
-
Is it possible to track objects on the go?
DeepSORT is one of the best trackers https://github.com/nwojke/deep_sortIt requires an object detector tho, like YOLO https://pjreddie.com/darknet/yolo/
What are some alternatives?
yolov5 - YOLOv5 🚀 in PyTorch > ONNX > CoreML > TFLite
sort - Simple, online, and realtime tracking of multiple objects in a video sequence.
yolor - implementation of paper - You Only Learn One Representation: Unified Network for Multiple Tasks (https://arxiv.org/abs/2105.04206)
yolov4-deepsort - Object tracking implemented with YOLOv4, DeepSort, and TensorFlow.
darknet - YOLOv4 / Scaled-YOLOv4 / YOLO - Neural Networks for Object Detection (Windows and Linux version of Darknet )
Similari - A framework for building high-performance real-time multiple object trackers
mmdetection - OpenMMLab Detection Toolbox and Benchmark
yolo_series_deepsort_pytorch - Deepsort with yolo series. This project support the existing yolo detection model algorithm (YOLOV8, YOLOV7, YOLOV6, YOLOV5, YOLOV4Scaled, YOLOV4, YOLOv3', PPYOLOE, YOLOR, YOLOX ).
YOLOv6 - YOLOv6: a single-stage object detection framework dedicated to industrial applications.