PAFNet: An Efficient Anchor-Free Object Detector Guidance

TL;DR

PAFNet is an efficient anchor-free object detection model that balances accuracy and speed for server and mobile applications, outperforming existing methods with a simplified structure and optimized tricks.

Contribution

The paper introduces PAFNet, a modified anchor-free detector based on TTFNet, optimized for both server and mobile environments, with improved efficiency and accuracy.

Findings

PAFNet achieves 42.2% mAP at 67.15 FPS on V100 GPU.

PAFNet-lite attains 23.9% mAP on Kirin 990 ARM CPU.

Outperforms existing state-of-the-art anchor-free detectors.

Abstract

Object detection is a basic but challenging task in computer vision, which plays a key role in a variety of industrial applications. However, object detectors based on deep learning usually require greater storage requirements and longer inference time, which hinders its practicality seriously. Therefore, a trade-off between effectiveness and efficiency is necessary in practical scenarios. Considering that without constraint of pre-defined anchors, anchor-free detectors can achieve acceptable accuracy and inference speed simultaneously. In this paper, we start from an anchor-free detector called TTFNet, modify the structure of TTFNet and introduce multiple existing tricks to realize effective server and mobile solutions respectively. Since all experiments in this paper are conducted based on PaddlePaddle, we call the model as PAFNet(Paddle Anchor Free Network). For server side, PAFNet can achieve a better balance between effectiveness (42.2% mAP) and efficiency (67.15 FPS) on a single V100 GPU. For moblie side, PAFNet-lite can achieve a better accuracy of (23.9% mAP) and 26.00 ms on Kirin 990 ARM CPU, outperforming the existing state-of-the-art anchor-free detectors by significant margins. Source code is at https://github.com/PaddlePaddle/PaddleDetection.