😄About Me

I am currently a mid-level multimodal algorithm engineer at the New Laboratory of Pattern Recognition(NLPR),Institute of Automation, Chinese Academy of Sciences(CASIA). I hold a bachelor’s degree in Electronic Information Science and Technology from Hebei Agricultural University and a master’s degree in Electronic and Communication Engineering from the School of Information and Communication Engineering, Beijing University of Posts and Telecommunications. My current work focuses on Pedestrian Recognition. I have a keen interest in research areas such as Embodied Intelligence and Fine-tuning of LLM

🚀Ongoing Work

Pedestrian Intention Prediction

Pedestrian intention prediction is a fundamental task in autonomous driving. Its purpose is to enable the vehicle to predict the crossing intentions of pedestrians on the road, allowing it to take appropriate actions, such as stopping to wait for pedestrians to cross or continuing to drive. Current pedestrian intention prediction generally uses pedestrian trajectories, surrounding environmental information, semantic information, pedestrian posture information, traffic light information, and other data to predict pedestrian intentions. Utilizing trajectory information and vehicle attribute information can achieve relatively good intention recognition performance. However, in complex real-world environments, relying solely on this information can still result in missed detections of pedestrian crossing intentions, thereby threatening pedestrian safety. We are dedicated to deeply exploring different modalities of data to reduce the missed detections of pedestrian crossing intentions, ensuring the safety of pedestrians on the road.

📘Areas of Interest

Planning to share some interesting work