Changhao Chen

I am a third-year Ph.D. student at Department of Computer Science, University of Oxford, under the supervision of Prof.Niki Trigoni and Prof. Andrew Markham, where I'm working on machine learning for signal processing, and intelligent sensor systems.

I am a member of Pembroke College, Oxford. I serve as the Vice President of Oxford China Forum 2019, and the Academic Officer of Pembroke MCR Committee.

Email: changhao.chen[at]cs.ox.ac.uk

Address: Robert Hooke Building, Parks Road, Oxford OX1 3PR, UK

Research

Driven by solving real-world problems, I'm applying representation learning, domain adaptation and multi-modalities learning for noisy sensor data with applictions on ubiquitous localization, robot navigation and robust perception & decisions.

I worked on Deep Learning based Inertial Odometry, including IONet, MotionTransformer, and OxIOD dataset.

I worked on Neural Visual-Inertial Navigation.

Selective Sensor Fusion for Neural Visual Inertial Odometry

Changhao Chen, Stefano Rosa, Yishu Miao, Chris Xiaoxuan Lu, Wei Wu, Andrew Markham, Niki Trigoni
CVPR-2019, Conference on Computer Vision and Pattern Recognition, Long Beach, USA, June 2019
[PDF (2.6 MB)] [Bibtex] [Project Website]

A general selective sensor fusion framework for monocular Visual-Inertial Odometry

Autonomous Learning for Face Recognition in the Wild via Ambient Wireless Cues

Chris Xiaoxuan Lu, Xuan Kan, Bowen Du, Changhao Chen Hongkai Wen, Andrew Markham, Niki Trigoni John A. Stankovic
WWW-2019, THE WEB CONFERENCE 2019, San Francisco, USA, May 2019
[PDF (1.9 MB)] [Bibtex]

AutoTune, which learns and refines the association between a face and wireless identifier over time, by increasing the inter-cluster separation and minimizing the intra cluster distance.



MotionTransformer: Transferring Neural Inertial Tracking Between Domains

Changhao Chen, Yishu Miao, Chris Xiaoxuan Lu, Linhai Xie, Phil Blunsom, Andrew Markham, Niki Trigoni
AAAI-2019, The 33rd AAAI Conference on Artificial Intelligence, Honolulu, USA, Feb 2019
[PDF (1.3 MB)] [Bibtex]

A robust generative adversarial network for sequence domain transformation, which is able to directly learn inertial tracking in unlabelled domains without using any paired sequences.

Deep Gait Recognition via Millimeter Wave

Peijun Zhao, Chris Xiaoxuan Lu, Jianan Wang, Changhao Chen Wei Wang, Andrew Markham, Niki Trigoni
EWSN-2019, The International Conference on Embedded Wireless Systems and Networks, Beijing, China, Feb 2019
[PDF (coming soon)] [Bibtex]

A human recognition system that identifies gaits based on millimeter wave (MMwave) using deep recurrent neural network.

Transferring Physical Motion Between Domains for Neural Inertial Tracking

Changhao Chen, Yishu Miao, Chris Xiaoxuan Lu, Phil Blunsom, Andrew Markham, Niki Trigoni
NeurIPS-2018, workshop on Modelling the Physical world: Perception, Learning and Control, Montreal, Canada, Dec 2018
[PDF (0.4 MB)] [Bibtex]

A short version of MotionTransformer

Learning with Stochastic Guidance for Navigation

Linhai Xie, Yishu Miao, Sen Wang, Phil Blunsom, Zhihua Wang, Changhao Chen, Andrew Markham, Niki Trigoni
NeurIPS-2018, Workshop on Probabilistic Reinforcement Learning and Structured Control, Montreal, Canada, Dec 2018
[PDF (2.3 MB)] [Bibtex]

A new framework for overcoming the high variance problem of DDPG by incorporating a stochastic switch, allowing an agent to choose between high and low variance policies.

OxIOD: The Dataset for Deep Inertial Odometry

Changhao Chen, Peijun Zhao, Chris Xiaoxuan Lu, Wei Wang, Andrew Markham, Niki Trigoni
arXiv:1809.07491, 2018
[PDF (2.4 MB)] [Bibtex]

The Oxford Inertial Odometry Dataset (OxIOD), a first-of-its-kind data collection for inertial-odometry research, contains 158 sequences totalling more than 42 km in total distance with all sequences having ground-truth labels.

IONet: Learning to Cure the Curse of Drift in Inertial Odometry

Changhao Chen, Chris Xiaoxuan Lu, Andrew Markham, Niki Trigoni
AAAI-2018, The 32nd AAAI Conference on Artificial Intelligence, New Orleans, USA, Feb 2018
[PDF (2.4 MB)] [Bibtex]

The first deep neural network (DNN) framework that learns location transforms in polar coordinates from raw IMU data, and constructs inertial odometry regardless of IMU attachment.

Simultaneous Localization and Mapping with Power Network Electromagnetic Field

Chris Xiaoxuan Lu, Yang Li, Peijun Zhao, Changhao Chen, Linhai Xie, Hongkai Wen, Rui Tan, Niki Trigoni
MobiCom-2018, Annual International Conference on Mobile Computing and Networking, New Delhi, India, Oct 2018
[PDF (4.0 MB)] [Bibtex]

This paper presents a first systematic study on using the electromagnetic field (EMF) induced by a building's electric power network for simultaneous localization and mapping (SLAM).

Course Projects
prl

Tightly-Coupled Integration of Inertial and Magneto-Inductive Sensors for Large-Scale Indoor Localization

Changhao Chen, Andrew Markham, Niki Trigoni
Reading Course Report, Oxford, UK, April 2017
[PDF (0.9 MB)] [Bibtex]

We presented Range-Constrained Kalman Filtering for hightly-coupled integration between IMU and MI sensors.

Teaching
pacman

Machine Learning - Michaelmas Term 2018 (Teaching Assistant)

Intelligent Systems - Michaelmas Term 2017 (Teaching Assistant)


This website was created from source code.