Amine Kacete
Papers
4
Total Citations
34
H-Index
4
About
Amine Kacete is a computer vision researcher specializing in 3D scene understanding, object pose estimation, and camera relocalization, with a particular focus on applications in augmented reality and robotics. His work bridges machine learning and geometric approaches to tackle some of the most persistent challenges in spatial computing. Kacete's most recognized contribution, L6DNet (2020), introduced a lightweight hybrid pipeline for 6 degrees-of-freedom (DoF) object pose estimation from single RGB-D images, demonstrating strong performance even with small training datasets — a significant practical advancement garnering 16 citations. His complementary work, YOLOff, further extended robust 6DoF estimation to cluttered real-world scenes using an offset-learning strategy. In the domain of camera relocalization, Kacete developed both a sparse feature regression forest method and the deep learning-based xyzNet, a scene coordinate prediction network designed to deliver the real-time accuracy that augmented reality systems demand. Each of these contributions, cited seven times respectively, reflects his consistent drive toward methods that are simultaneously fast, accurate, and deployable. Across his portfolio, Kacete demonstrates a talent for designing pragmatic hybrid systems that combine the strengths of data-driven learning with principled geometric reasoning — an approach that positions his research as highly relevant to next-generation AR and autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4YOLOff: You Only Learn Offsets for robust 6DoF object pose estimation4 citations · 2020
Key Collaborators
Related papers
- L6DNet: Light 6 DoF Network for Robust and Precise Object Pose Estimation with Small Datasets
- L6DNet: Light 6 DoF Network for Robust and Precise Object Pose Estimation with Small Datasets
- YOLOff: You Only Learn Offsets for robust 6DoF object pose estimation
- xyzNet: Towards Machine Learning Camera Relocalization by Using a Scene Coordinate Prediction Network
- Robust 6D Object Pose Estimation by Learning RGB-D Features
Researchers in this area
Labs working in this area
If you're exploring the applied side
Commercial systems in adjacent areas you may find useful as a starting point. This is not a claim that the research here is used in them — academia and industry often move on separate tracks.
- Related commercial systems
Alphasense PositionSevensense
Suggested by topic similarity — not advertising or endorsement.