Papers
2
Total Citations
6
H-Index
2
About
Alina Marcu is a computer vision researcher whose work focuses on semantic segmentation, aerial and drone video analysis, and automated annotation methods — areas with significant implications for autonomous navigation and robotic safety systems. Her research addresses one of the most persistent challenges in deep learning: the labor-intensive process of generating pixel-level annotations required to train accurate segmentation models. In her notable 2019 work, "Towards Automatic Annotation for Semantic Segmentation in Drone Videos," Marcu explored ways to reduce the annotation burden that has long slowed progress in aerial imagery research, recognizing that ground-level dataset advances had not translated easily to drone perspectives. Building on this foundation, her 2021 paper, "Semantics Through Time: Semi-supervised Segmentation of Aerial Videos with Iterative Label Propagation," proposed a semi-supervised approach that leverages temporal consistency across video frames to propagate labels iteratively — reducing the need for exhaustive manual annotation while maintaining segmentation quality. Both papers have garnered early citations, signaling growing community interest in her methods. Marcu's contributions are particularly relevant as drone-based perception becomes increasingly critical in autonomous systems, smart infrastructure monitoring, and search-and-rescue applications, positioning her as an emerging voice in efficient, scalable computer vision for aerial domains.
Research Focus
Key Achievements
Top Papers
- 1
- 2Towards Automatic Annotation for Semantic Segmentation in Drone Videos3 citations · 2019