← Search

Fengji Ma

2 accepted papers

2026

Cueing Without Gapping: Cuer-Independent Cued Speech Recognition Powered by Cross-Cuer Invariant Modeling

AAAI 2026technical

Automatic Cued Speech Recognition (ACSR) is a vital communication system designed to enhance spoken language accessibility for the hearing-impaired by combining lip movements and hand gestures to encode phonemes. Despite its effectiveness, current ACSR methods face significant challenges, including

Cited by 0SourcePDFScholar
2026

TIMA: Text-Image Mutual Awareness for Balancing Zero-Shot Adversarial Robustness and Generalization Ability

AAAI 2026technical

Achieving zero-shot adversarial robustness without sacrificing generalization remains challenging for foundation models such as CLIP, especially under large adversarial perturbations. Through empirical analyses, we identify three critical yet overlooked issues: (1) Logit margins exhibit a stable off

Cited by 0SourcePDFScholar