Ygor C. N. Sousa

Papers

Total Citations

H-Index

About

Ygor C. N. Sousa is a researcher advancing the field of robotic perception and semantic mapping. His work focuses on enabling autonomous agents to build rich, topological representations of their environments by integrating deep visual features and unsupervised learning. Sousa’s most notable contribution, “Topological Semantic Mapping by Consolidation of Deep Visual Features” (2022), has garnered 16 citations for its novel approach to embedding semantic properties—such as room categories and object classes—into spatial maps using convolutional neural networks. This work bridges the gap between raw visual data and high-level environmental understanding, a critical step for intelligent navigation. Earlier, in “Incremental Semantic Mapping with Unsupervised On-line Learning” (2018), he pioneered an on-line, unsupervised method using Self-Organizing Maps (SOM) to incrementally build topological maps enriched with object recognition, laying groundwork for adaptive, lifelong learning in robotics. Sousa’s research is impactful for students and engineers developing autonomous systems that must interpret and navigate complex, dynamic spaces. His achievements highlight a commitment to scalable, data-efficient mapping solutions that push the boundaries of robotic cognition.