Maxim Kazakov

Alexey Sidnev, Ekaterina Krasikova, Maxim Kazakov

The success of deep neural networks in the traditional keypoint detection task encourages researchers to solve new problems and collect more complex datasets. The size of the DeepFashion2 dataset poses a new challenge on the keypoint detection task, as it comprises 13 clothing categories that span a wide range of keypoints (294 in total). The direct prediction of all keypoints leads to huge memory consumption, slow training, and a slow inference time. This paper studies the keypoint grouping approach and how it affects the performance of the CenterNet architecture. We propose a simple and efficient automatic grouping technique with a powerful post-processing method and apply it to the DeepFashion2 fashion landmark task and the MS COCO pose estimation task. This reduces memory consumption and processing time during inference by up to 19% and 30% respectively, and during the training stage by 28% and 26% respectively, without compromising accuracy.

Alexey Sidnev, Alexander Krapivin, Alexey Trushkov et al.

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.

Alexey Sidnev, Alexey Trushkov, Maxim Kazakov et al.

The one-shot approach, DeepMark, for fast clothing detection as a modification of a multi-target network, CenterNet, is proposed in the paper. The state-of-the-art accuracy of 0.723 mAP for bounding box detection task and 0.532 mAP for landmark detection task on the DeepFashion2 Challenge dataset were achieved. The proposed architecture can be used effectively on the low-power devices.