Ruiqi Xian (先睿奇)

I'm a third-year PhD in Electrical and Computer Engineering at the University of Maryland College Park, where I am involved in advanced research under the guidance of Dr. Dinesh Manocha at the GAMMA Lab. My primary focus lies in the realms of computer vision and robotics, with a specialization in video processing and understanding.

Currently, I am working on perception problems from videos captrued by Unmanned Aerial Vehicles(UAVs). Although my research is primarily centered on aerial scene perception, I am also very interested in topics related to Video Foundation Models and Self-Supervised Learning.

Email  /  CV  /  Scholar  /  Twitter  /  Github

profile photo
News
  • (Nov 2023) HallusionBench is on ArXiv!
  • (OCT 2023) MITFAS and PMI Sampler have been accepted to WACV 2024.
  • (Sep 2023) PLAR is on Arxiv!
  • (Jan 2023) AZTR has been accepted to ICRA 2023.
Publications
Aerial_Booth HallusionBench: An Advanced Diagnostic Suite for Entangled Language Hallucination & Visual Illusion in Large Vision-Language Models
Tianrui Guan*, Fuxiao Liu*, Xiyang Wu, Ruiqi Xian, Zongxia Li, Xiaoyu Liu, Xijun Wang, Lichang Chen, Furong Huang, Yaser Yacoob, Dinesh Manocha, Tianyi Zhou
ArXiv

An comprehensive benchmark designed for the evaluation of image-context reasoning, which presents significant challenges to advanced large visual-language models (LVLMs), such as GPT-4V(Vision) and LLaVA-1.5, by emphasizing nuanced understanding and interpretation of visual data.

Aerial_Diffusion PLAR: Prompt Learning for Action Recognition
Ruiqi Xian*, Xijun Wang*, Tianrui Guan, Dinesh Manocha
ArXiv

A novel action recognition approach that leverages the strengths of prompt learning to optimize the learning process.

DifFAR PMI Sampler: Patch Similarity Guided Frame Selection for Aerial Action Recognition
Ruiqi Xian, Xijun Wang, Divya Kothandaraman, Dinesh Manocha
WACV 2024

A frame selection strategy utilizes the motion bias within videos via patch-wise similarity, enabling the selection of motion-salient frames with dynamic background.

SALAD MITFAS: Mutual Information based Temporal Feature Alignment and Sampling for Aerial Video Action Recognition
Ruiqi Xian*, Xijun Wang, Dinesh Manocha
WACV 2024

A novel alignment and sampling approach that roots in information theory to handle the viewpoint changes caused by the UAV ego-motions.

FAR AZTR: Aerial Video Action Recognition with Auto Zoom and Temporal Reasoning
Ruiqi Xian*, Xijun Wang*, Tianrui Guan, Celso M. de Melo, Stephen M. Nogar, Aniket Bera, Dinesh Manocha
ICRA 2023

A learning-based approach that can be implemented and evaluated both on the desktop with high-end GPUs and on the low power Platforms for robots and drones.


Website template borrowed from Jon Barron.