Hang Yang
PhD candidate, Department of Mathematics & Statistics, University of Massachusetts Amherst.
hangyang@umass.edu · Google Scholar · GitHub
Research
I work on machine learning, invertible neural networks, and robotics.
- Mobile motion capture. Using a quadruped robot as a mobile platform to record human motion outside the constraints of a fixed studio.
- Invertible neural networks for lossless information recovery, applied to deep-learning image steganography (PRIS, DKiS).
- Medical imaging. Vision transformers for automated disease detection — CovidViT reaches 98% on chest X-rays.
- Robotics. Legged-robot perception, autonomy, and human–robot interaction.
Publications
-
2025
DKiS: Decay weight invertible image steganography with private key
Introduces a decay weight mechanism and private key integration into invertible image steganography, enhancing both steganographic performance and security.
-
2024
PRIS: Practical robust invertible network for image steganography
A practical robust invertible network for image steganography that achieves state-of-the-art robustness against multiple distortion attacks using pre/post-enhancement modules and a gradient approximation function.
-
2022
CovidViT: a novel neural network with self-attention mechanism to detect Covid-19 through X-ray images
First application of a vision transformer with self-attention mechanism to COVID-19 detection from chest X-rays, achieving 98.0% accuracy on the test set.
Teaching
-
2024
MATH 233: Calculus 3
Teaching assistant for MATH 233: Calculus 3 at UMass Amherst.
-
2023
MATH 132: Calculus 2
Teaching assistant for MATH 132: Calculus 2 at UMass Amherst.
Projects
-
2026
AmazingHand: Dexterous Robotic Hand Build
A personal build of the open-source AmazingHand — a 3D-printed dexterous robotic hand with independently actuated fingers.
-
2026
MWIP — invertible-network image watermarking
A 4M-param INN that embeds an 896-bit cryptographic provenance payload (perceptual hash + Ed25519 signature + public key) directly into image pixels — survives JPEG re-encoding and benign resize down to social-media chain levels.
Education
- 2024 — present
- Ph.D., Mathematics, University of Massachusetts Amherst
- 2021 — 2024
- M.Sc., Mathematics, China Agricultural University
- 2017 — 2021
- B.Sc., Mathematics, China Agricultural University