Sichao Tian

National University of Singapore, Chinese Academy of Medical Sciences & Peking Union Medical College

Papers

Total Citations

H-Index

About

Sichao Tian is a robotics researcher advancing the frontier of few-shot and incremental learning for object detection. Their work focuses on enabling robots to learn new visual tasks quickly from minimal examples while retaining prior knowledge—a critical capability for autonomous systems operating in dynamic environments. Tian’s foundational paper, “Incremental Few-Shot Object Detection for Robotics” (2022), has garnered 15 citations for addressing the dual challenge of data efficiency and continuous learning in real-world robotics. Building on this, their 2024 study, “Bilateral-Head Region-Based Convolutional Neural Networks,” introduces a unified framework that integrates incremental and few-shot detection, achieving robust performance on evolving datasets with limited supervision. This work has direct implications for autonomous driving and robotics, where large-scale annotated data is often impractical. Tian’s contributions are shaping the next generation of open-ended perception systems, allowing robots to adapt flexibly without catastrophic forgetting. Their research stands at the intersection of computer vision and robotics, offering scalable solutions for lifelong learning in resource-constrained environments.