Research

My work spans small object detection, visual representation learning, and efficient vision-language models, with recent work on adaptive resolution for efficient vision-language understanding.

set figure2

SET: Spectral Enhancement for Tiny Object Detection

Huixin Sun, Runqi Wang, Yanjing Li, Linlin Yang, Shaohui Lin, Xianbin Cao, Baochang Zhang

CVPR 2025

SET improves tiny object detection by suppressing high-frequency background noise and using adversarial perturbations during training to strengthen object features.

P4Q Figure 1: image–text cosine similarity predictions

P4Q: Learning to Prompt for Quantization in Low-Bit CLIP

Huixin Sun, Runqi Wang, Yanjing Li, Chang Gao, Liping Jing, Xiaolong Jiang, Yao Hu, Baochang Zhang, Xianbin Cao

ACM TOMM

P4Q adapts post-training quantized CLIP to downstream tasks through learnable prompts, low-bit adapters for image–text feature alignment, and distillation from a full-precision teacher.

tolf figure3

Noise-Robust Tiny Object Localization with Flows

Huixin Sun, Linlin Yang, Rongyu Chen, Kerui Gu, Baochang Zhang, Angela Yao, Xianbin Cao

Pattern Recognition

We improve robustness to noisy annotations in tiny object localization by modeling localization errors with normalizing flows and downweighting gradients from high-uncertainty samples.

q yolo figure1

Q-YOLO: Efficient Inference for Real-Time Object Detection

Mingze Wang*, Huixin Sun*, Jun Shi*, Xuhui Liu, Xianbin Cao, Luping Zhang, Baochang Zhang

ACPR 2023

Q-YOLO addresses accuracy degradation in post-training quantized YOLO detectors by calibrating activation clipping thresholds to minimize quantization error under imbalanced activation distributions.

* Equal contribution.