DeepMark++: Real-time Clothing Detection at the Edge

TL;DR

DeepMark++ presents a fast, efficient clothing detection model suitable for real-time edge devices, achieving high accuracy and speed on mobile hardware, advancing fashion AI applications.

Contribution

The paper introduces a single-stage multi-target network with novel post-processing for real-time clothing detection and keypoint estimation on edge devices.

Findings

Achieves comparable accuracy to state-of-the-art on DeepFashion2

Runs at 17 FPS on Huawei P40 Pro

Secures second place in DeepFashion2 Landmark Estimation Challenge 2020

Abstract

Clothing recognition is the most fundamental AI application challenge within the fashion domain. While existing solutions offer decent recognition accuracy, they are generally slow and require significant computational resources. In this paper we propose a single-stage approach to overcome this obstacle and deliver rapid clothing detection and keypoint estimation. Our solution is based on a multi-target network CenterNet, and we introduce several powerful post-processing techniques to enhance performance. Our most accurate model achieves results comparable to state-of-the-art solutions on the DeepFashion2 dataset, and our light and fast model runs at 17 FPS on the Huawei P40 Pro smartphone. In addition, we achieved second place in the DeepFashion2 Landmark Estimation Challenge 2020 with 0.582 mAP on the test dataset.