# hunglc007/tensorflow-yolov4-tflite

YOLOv4, YOLOv4-tiny, YOLOv3, YOLOv3-tiny Implemented in Tensorflow 2.0, Android. Convert YOLO v4 .weights tensorflow, tensorrt and tflite

Repository: https://github.com/hunglc007/tensorflow-yolov4-tflite
Canonical: https://ross.abutalabs.com/products/tensorflow-yolov4-tflite
Homepage: https://github.com/hunglc007/tensorflow-yolov4-tflite
Language: Python
License: MIT
License Family: permissive
Topics: yolov4, yolov3, tflite, object-detection, tensorflow, tf2, tensorrt, yolov3-tiny, android
Last push: 2024-05-12T16:19:31+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2322, "days_push": 843, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2254, forks 1220 (observed 2026-08-28T04:06:31.219611+00:00)

## What it is
A TensorFlow 2.x implementation of YOLOv4, YOLOv4-tiny, YOLOv3, and YOLOv3-tiny object detection models, with scripts that convert original Darknet .weights into TensorFlow, TensorFlow Lite, and TensorRT formats. It includes detection and video demos, training, mAP evaluation, and float16/int8 quantization for deployment on desktop, NVIDIA GPUs, and Android.

## Use cases
- convert darknet yolov4 weights to tensorflow 2
- run yolov4 object detection on images or video in python
- convert yolo model to tflite for android deployment
- quantize yolov4 to int8 or float16 tflite
- convert yolo weights to tensorrt for faster gpu inference
- train yolov4 on a custom dataset in tensorflow
- benchmark yolov4-tiny for real-time detection

## When to choose
- You need official Darknet YOLOv4/v3 weights running inside a TensorFlow 2.x pipeline
- You are deploying to mobile or edge devices and need TFLite conversion with quantization
- You want TensorRT-optimized YOLO inference on NVIDIA GPUs
- You want reference code for training or evaluating YOLO models in TensorFlow rather than Darknet

## When to avoid
- You want newer YOLO generations (YOLOv7/v8+) or prefer the PyTorch ecosystem
- You need production-grade int8 quantized YOLOv4/v4-tiny, which the README acknowledges has known issues
- Your stack requires a modern TensorFlow version, since the project targets TF 2.3-era APIs
- You need a maintained turnkey detection service rather than scripts and model code

## Facets
- artifact type: library
- maturity: maintenance
- function: deep-learning, machine-learning, computer-vision, image-processing, gpu-computing
- domain: computer-vision, deep-learning, machine-learning, mobile-development
- platform: python, cross-platform
- tags: yolov4, yolov3, object-detection, tensorflow2, tflite, tensorrt, model-conversion, darknet-weights, int8-quantization, real-time-detection, android, gpu

## Member repositories
- hunglc007/tensorflow-yolov4-tflite (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:31.219611+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:43:46.346700+00:00, confidence not recorded.
  - readme: https://github.com/hunglc007/tensorflow-yolov4-tflite (fetched 2026-08-28T04:06:31.219611+00:00, sha 398f7caef7e7)
  - homepage: https://github.com/hunglc007/tensorflow-yolov4-tflite (fetched 2026-08-29T10:23:51.612298+00:00, sha 92906c354261)
- Data as of 2026-08-30T08:39:29.467469+00:00.