Practical overview
When it fits
Use the repository when you need a TensorFlow 2 implementation with pretrained YOLOv3 variants and code examples for detection, training or model export.
Who it is for
- Developers exploring object detection with TensorFlow 2.
- Practitioners adapting a YOLOv3 detector to custom data.
- Teams that want to export the model for TensorFlow Serving.
Documented uses
- Detect objects in an image, webcam stream or video file.
- Fine-tune or train a detector using TFRecord-formatted data.
- Compare eager-mode and graph-mode training examples.
- Export the model as a TensorFlow Serving SavedModel.
Requirements & limits
The repository is public and identifies its license as MIT. The README documents Conda CPU and GPU environment files and pip installation; GPU use requires an NVIDIA driver.
- The repository includes both YOLOv3 and YOLOv3-tiny pretrained-weight workflows.
- Training examples use eager mode with tf.GradientTape and graph-mode fitting with model.fit.
- Custom training expects TFRecord data formatted for the TensorFlow Object Detection API.
- The Serving export expects preprocessed images and returns bounding boxes, scores, classes and valid-detection counts.
- The README's benchmark figures are described as rough calculations, not guarantees for other configurations.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Run detection on an image
Set up the repository and use a converted pretrained YOLOv3 model with the documented image-detection example.
Before you start
- A local copy of the repository.
- Conda with the repository's CPU or GPU environment file, or pip to install requirements.
- Pretrained Darknet YOLOv3 weights.
Steps
Create and activate the CPU or GPU Conda environment, or install dependencies with pip.
Download the YOLOv3 Darknet weights to data/yolov3.weights.
Run the documented conversion command to create checkpoints/yolov3.tf.
Run the image detection example with python detect.py --image ./data/meme.jpg.
Expected outcome
The documented workflow runs the YOLOv3 detector on the specified image.
Guide limitations
- The GPU setup also requires an NVIDIA driver.
- The repository supplies the workflow; this guide does not establish performance on a particular machine.
Compare options for Model inference and serving
These projects document a common task and share a project role. This is a source-based comparison, not an execution benchmark or a guarantee of interchangeability.
| Project | Documented purpose | Platform | Access and requirements |
|---|---|---|---|
| YOLOV3 TF2 | Detect objects in an image, webcam stream or video file. | Cloud | The repository is public and identifies its license as MIT. The README documents Conda CPU and GPU environment files and pip installation; GPU use requires an NVIDIA driver. |
| Pulse Of Motion | Estimate a video's physical frame rate from its visual dynamics. | Linux | The repository is public. Its README describes a released checkpoint that downloads from Hugging Face on first inference, with a manual download option. The strongest internal model is not fully released; the README directs interested users to contact the project about access. No license information is provided in the repository evidence. |
| Neural Image Assessment | Train the model using AVA images and rating annotations. | Cloud, Linux, macOS, Windows |
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 14 October 2026
- Original website title
- zzh8829/yolov3-tf2 · Official homepage title · retained in its source language