基於人體步態序列之性別分類與年齡估計方法的相關研究

A Study of Gait-Based Gender Classification and Age Estimation Using Pose Sequences

楊日翔(Jih-Hsiang Yang)

指導教授 ： 吳家麟

國立臺灣大學/電機資訊學院/資訊工程學研究所/碩士(2021年)

https://doi.org/10.6342/NTU202201964

摘要

本篇論文將時空圖卷積網路應用在基於人體姿態序列的步態資料集，進而進行性別分類和年齡估計，這兩項工作即時使用人眼作為判斷依據，仍然可能產生很大的誤差。透過一連串實驗，我們發現，適當的結合多任務學習和資料擴增的方法，可以進一步提升性別分類和年齡估計的正確率。

關鍵字

人類步態 ； 姿態序列 ； 性別分類 ； 年齡估計 ； 圖神經網路 ； 多任務學習

並列摘要

This work presents a Spatial-Temporal Graph Convolutional Network (ST-GCN) for gender classification and age estimation using pose sequences of gait, which may be hard to comprehend by human eyes. A series of experiments show that an appropriate combination of multi-task learning and data augmentation does improve the expected performances.

並列關鍵字

Human Gait ； Pose Sequences ； Gender Classification ； Age Estimation ； Graph Neural Networks ； Multi-Task Learning

參考文獻

[1] Jinguang Han and Bir Bhanu. Individual recognition using gait energy image. IEEE transactions on pattern analysis and machine intelligence, 28(2):316–322, 2005.

[2] Weizhi An, Shiqi Yu, Yasushi Makihara, Xinhui Wu, Chi Xu, Yang Yu, Rijun Liao, and Yasushi Yagi. Performance evaluation of model-based gait on multi-view very large population database with pose sequences. IEEE Transactions on Biometrics, Behavior, and Identity Science, 2(4):421–430, 2020.

[3] Martin Hofmann, Shamik Sural, and Gerhard Rigoll. Gait recognition in the presence of occlusion: A new dataset and baseline algorithms. 2011.

[4] Shiqi Yu, Daoliang Tan, and Tieniu Tan. A framework for evaluating the effect of view angle, clothing and carrying condition on gait recognition. In 18th International Conference on Pattern Recognition (ICPR’06), volume 4, pages 441–444. IEEE, 2006.

[5] Zhe Cao, Tomas Simon, Shih-En Wei, and Yaser Sheikh. Realtime multi-person 2d pose estimation using part affinity fields. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 7291–7299, 2017.