Shichao Wu, Assistant Professor at Southwest University of Science and Technology

Shichao Wu is an Assistant Professor at Southwest University of Science and Technology and a Researcher at the Robot Technology Used for Special Environment Key Laboratory of Sichuan Province, School of Information and Control Engineering. His primary research areas include acoustic computing and spatial perception intelligence. He has published a total of 30 academic papers, including 18 SCI papers, 13 top journal papers, and 1 highly cited paper. Previously, he obtained a B.E. from Southwest University of Science and Technology in 2017, M.S. from Northeastern University in 2020, and Ph.D. in Artificial Intelligence from Nankai University in 2024, under the supervision of Prof. Jingtai Liu.

吴仕超，西南科技大学特聘副教授，信息工程学院特殊环境机器人技术四川省重点实验室研究人员。 主要研究方向为声学计算、空间感知智能，已发表学术论文共30篇，其中SCI期刊论文18篇，TOP期刊论文13篇，高被引论文1篇。 在此之前，他于2017年在西南科技大学取得学士学位，2020年在东北大学取得硕士学位， 2024年在南开大学取得博士学位（人工智能专业，导师：刘景泰教授）。

Email  /  Google Scholar  /  CSDN  /  Github

As the Principal Investigator (PI), I have led multiple research projects, including one Natural Science Foundation of Sichuan Province, one National Natural Science Foundation of China (NSFC) Youth Fund, and one Outstanding Talent Cultivation Fund Project at Southwest University of Science and Technology. In addition, I have participated one Natural Science Foundation of Sichuan Province (General Program) and one NSFC General Program.

I’m currently engaged in and plan to research spatial perception intelligence for robotics. These works primarily focus on (a) acoustic computing; (b) audio-visual perception; and (c) spatial perception intelligence-enabled dexterous manipulation and embodied navigation. These works tried to perceive and understand the spatial characteristics of humans and objects from both visual and auditory perspectives, thereby enabling Generalist Robots to operate (navigate and manipulate) effectively in various complex scenarios.

My publications can be found here. See also: Google Scholar. Here are some recent publications. Representative papers are highlighted.

We design a Dominant Frequency Segmented Conformer with Domain Adaptation (DFSC-DA) to enhance acoustic footstep-based person identification under interferences (clothing, shoe type, room type). It segments raw footsteps by dominant frequencies, models intra/inter-period variations via Conformer, and uses adversarial training for robust features.

We propose a contrastive representation-based multimodal acoustic model (CRATI) for sound event localization and detection (SELD), which learns robust audio representations from audio, text, and image in an end-to-end manner.

We build the acoustic footstep-based person identification and localization dataset (AFPILD) by unifying the identify and locate tasks for the first time, concerning the clothing and shoe type covariates.

We propose one hierarchical audio augmentation chain (HAAC) that contains feature map augmentation, audio channel swapping, and sample mixup to augment the audio features for the SELD task.

We propose one improved acoustic footstep-based person identification dataset (AFPID-II) from 41 subjects, counting over 14 h of footstep audios.

We propose the hierarchical contexts (the entity context, the global context, and the scene context) based emotion recognition method with scene graphs.

Propose one driving fatigue detection approach based on the brain functional network, and provide one way for the key EEG electrodes and rhythms selection.

Propose the EFDMs based on EEG with STFT, and combine them with the deep convolutional neural network for emotion recognition.

Propose one driving fatigue detection method based on multiple nonlinear features fused with multiple kernel learning (MKL) based SVM.

Propose one fatigue detection approch with transfer learning based on the Electrode-Frequency Distribution Maps (EFDMs) of EEG signals.

"Highly Cited Paper of 2019" in the Journal of Electronics & Information Technology

Early-Career Editorial Board of Robot Learning

Reviewer of IEEE Transactions on Neural Networks and Learning Systems (TNNLS), Engineering Applications of Artificial Intelligence (EAAI), International Journal of Intelligent Systems, Applied Intelligence, IROS, et al.

Special thanks to Jon Barron for the website's source code

Last updated: Oct. 2025