|11 months ago||6 days ago|
|-||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.
We haven't tracked posts mentioning human-detection yet.
Tracking mentions began in Dec 2020.
Roboflow 100: A New Object Detection Benchmark
5 projects | news.ycombinator.com | 28 Dec 2022
Haven't heard of those two, but would be really awesome to see an integration. We have an open API for just this reason: we really want to make it easy to use (and source) your data across all the different tools out there. We've recently launched integrations with other labeling and AutoML tools (and have integrations with the big-cloud AutoML tools as well). We're hoping to have a bunch more integrations with other MLOps tools & platforms in 2023.
Re synthetic data specifically, we've written a couple of how-to guides for creating data from context augmentation, Unity Perception, and Stable Diffusion & are talking to some others as well; it seems like a natural integration point (and someplace where we don't need to reinvent the wheel).
Football Players Tracking with YOLOv5 + ByteTRACK Tutorial
3 projects | reddit.com/r/computervision | 7 Dec 2022
I love this! TF.js is big! And to answer your question - sure, we can run that model in the browser. This is the YOLOv5 model it can be converted from PyTorch to TF.js with this script: https://github.com/ultralytics/yolov5/blob/master/export.py And then run it with my NPM package https://github.com/SkalskiP/yolov5js. ML in Java Script is the future! The problem is I don't know anything about any good tracker implemented in JS.
Need to download resources for DeepSORT from pan.baidu.com
3 projects | reddit.com/r/computervision | 6 Dec 2022
The yolov5 models are available at ultralytics but I am guessing you need to .onnx it first (the site is not in english for me so I can't read it).
What is Pruning YOLO?
2 projects | reddit.com/r/computervision | 4 Nov 2022
Here is the pruning yolov5, did by the author. https://github.com/ultralytics/yolov5/issues/304
Use YOLOv5 tensorflow.js models to speed up annotation
2 projects | reddit.com/r/computervision | 3 Oct 2022
Hi everyone! I'm Piotr and for several years I have been developing a small open-source project for labeling photos - makesense.ai. I added a new feature this weekend. You can use [YOLOv5](https://github.com/ultralytics/yolov5) models to automatically annotate photos. You can choose one of the models pre-trained on the COCO dataset, but most importantly you can load your own custom models. Just drag and drop the tensorflow.js model to the editor and you are good to go. Everything runs in brawser - no backend, so it is completely free. Let me know what you think! I'm super excited about that project.
DeepSort with PyTorch(support yolo series)
13 projects | reddit.com/r/u_No_Experience9104 | 20 Sep 2022
Build Custom Functions for YOLOv4 with TensorFlow, TFLite & TensorRT
2 projects | reddit.com/r/tensorflow | 3 Aug 2022
Is there a reason to use YOLOv4 over YOLOv5 or YOLOR?
Data augmentation PyTorch transforms for object detection data with bounding boxes
2 projects | reddit.com/r/computervision | 26 Jul 2022
The other common one is Albumentations and many modern object detection models perform online augmentation during the training process.
What are the fastest face decection liblaries for detect faces in a video?
3 projects | reddit.com/r/computervision | 11 Jul 2022
https://github.com/ultralytics/yolov5 (scroll down to Inference with detect.py in the ReadMe)
Inconsistent detections after custom training with YOLOv5
4 projects | reddit.com/r/computervision | 4 Jul 2022
I included the training metrics in a Github issue. Training and validation are executed on the same type of images (real X-rays with simulated wrenches), so the train/val metrics are pretty good.
What are some alternatives?
detectron2 - Detectron2 is a platform for object detection, segmentation and other visual recognition tasks.
darknet - YOLOv4 / Scaled-YOLOv4 / YOLO - Neural Networks for Object Detection (Windows and Linux version of Darknet )
mmdetection - OpenMMLab Detection Toolbox and Benchmark
yolor - implementation of paper - You Only Learn One Representation: Unified Network for Multiple Tasks (https://arxiv.org/abs/2105.04206)
Deep-SORT-YOLOv4 - People detection and optional tracking with Tensorflow backend.
yolov3 - YOLOv3 in PyTorch > ONNX > CoreML > TFLite
OpenCV - Open Source Computer Vision Library
edge-tpu-tiny-yolo - Run Tiny YOLO-v3 on Google's Edge TPU USB Accelerator.
yolov5-crowdhuman - Head and Person detection using yolov5. Detection from crowd.
CenterNet - Object detection, 3D detection, and pose estimation using center point detection:
YOLOv6 - YOLOv6: a single-stage object detection framework dedicated to industrial applications.
Mask-Detection-YOLOv3 - Mask Detection with YOLOv3.