Hierarchical binary CNNs for landmark localization with limited resources

Bulat, Adrian; Tzimiropoulos, Georgios

Abstract

Our goal is to design architectures that retain the groundbreaking performance of Convolutional Neural Networks (CNNs) for landmark localization and at the same time are lightweight, compact and suitable for applications with limited computational resources. To this end, we make the following contributions: (a) we are the first to study the effect of neural network binarization on localization tasks, namely human pose estimation and face alignment. We exhaustively evaluate various design choices, identify performance bottlenecks, and more importantly propose multiple orthogonal ways to boost performance. (b) Based on our analysis, we propose a novel hierarchical, parallel and multi-scale residual architecture that yields large performance improvement over the standard bottleneck block while having the same number of parameters, thus bridging the gap between the original network and its binarized counterpart. (c) We perform a large number of ablation studies that shed light on the properties and the performance of the proposed block. (d) We present results for experiments on the most challenging datasets for human pose estimation and face alignment, reporting in many cases state-of-the-art performance. (e) We further provide additional results for the problem of facial part segmentation. Code can be downloaded from https://www.adrianbulat.com/binary-cnn-landmarks

Citation

Bulat, A., & Tzimiropoulos, G. (2020). Hierarchical binary CNNs for landmark localization with limited resources. IEEE Transactions on Pattern Analysis and Machine Intelligence, 42(2), 343 - 356. https://doi.org/10.1109/tpami.2018.2866051

• Journal Article Type: Article
• Acceptance Date: Aug 5, 2018
• Online Publication Date: Aug 23, 2018
• Publication Date: 2020-02
• Deposit Date: Aug 16, 2018
• Publicly Available Date: Sep 11, 2018
• Journal: IEEE Transactions on Pattern Analysis and Machine Intelligence
• Print ISSN: 0162-8828
• Electronic ISSN: 1939-3539
• Publisher: Institute of Electrical and Electronics Engineers
• Peer Reviewed: Peer Reviewed
• Volume: 42
• Issue: 2
• Pages: 343 - 356
• DOI: https://doi.org/10.1109/tpami.2018.2866051
• Keywords: Binary Convolutional Neural Networks; Residual learning; Landmark localization; Human pose estimation; Face alignment.
• Public URL: https://nottingham-repository.worktribe.com/output/1036353
• Publisher URL: https://ieeexplore.ieee.org/document/8444745/
• Contract Date: Sep 11, 2018