Research & Publications

We advance the field through rigorous research, reproducible experiments, and open publication of our findings. Our work focuses on practical innovations that push state-of-the-art benchmarks.

Recent Publications

Research Areas

Our research spans multiple domains with a focus on practical impact and reproducibility.

🔬

Model Efficiency

Developing techniques for model compression, quantization, and efficient architectures that maintain SOTA performance with reduced computational requirements.

  • Knowledge distillation
  • Neural architecture search
  • Pruning & quantization
  • Efficient attention mechanisms
🏥

Medical Imaging

Advancing diagnostic AI with focus on segmentation, detection, and classification tasks across multiple imaging modalities.

  • Multi-modal learning
  • Few-shot medical AI
  • Uncertainty quantification
  • Explainable predictions
👁️

Computer Vision

Pushing boundaries in visual perception for real-world deployment scenarios including autonomous systems and industrial applications.

  • Real-time object detection
  • Video understanding
  • 3D scene reconstruction
  • Multi-task learning
📊

Benchmarking & Evaluation

Creating standardized evaluation protocols and comprehensive benchmarks that enable fair comparison of ML methods.

  • Reproducibility frameworks
  • Dataset curation
  • Evaluation metrics
  • Leaderboard infrastructure

Preprints & Working Papers

Talks & Presentations

Collaborations

We actively collaborate with academic institutions and research labs. Interested in working together?

Get in touch Email directly