Nicu Sebe
Papers
11
Total Citations
299
H-Index
8
About
Nicu Sebe is a leading figure at the intersection of computer vision, human-computer interaction, and deep learning, with a career dedicated to teaching machines to perceive, understand, and interact with the world. His foundational work on monocular depth estimation—a paper amassing over 130 citations—introduced a novel deep model using multi-scale continuous CRFs, significantly advancing how single images can infer 3D structure for robotics and augmented reality. Sebe has also been a pioneer in indoor localization, developing multi-view and thermal imaging techniques that enable devices to determine their position without relying on GPS or WiFi, a critical contribution for autonomous navigation in complex environments. His research extends to emotion recognition, where he has explored unsupervised pre-training to maintain accuracy even when faces are masked, and to object pose estimation, with recent diffusion-based methods achieving domain generalization. With a career spanning seminal surveys and highly cited papers, Sebe’s work has shaped modern vision systems, earning him recognition as a transformative researcher whose innovations bridge perception and real-world application.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4Indoor localization via multi-view images and videos21 citations · 2017
- 5
- 6Deep Learning-Based Object Pose Estimation: A Comprehensive Survey16 citations · 2026
- 7
- 8
- 9The age of human computer interaction8 citations · 2007
- 10Multi-Stage Multimodal Distillation for Audio-Visual Speaker Tracking4 citations · 2025