01
Learning-based light-field compression
Doctoral research on hybrid and learning-based approaches to light-field coding, including lossy, lossless, and distributed compression methods.
Computer vision · Image processing · Machine learning
I build reliable vision systems, from image data to deployment.
I am Milan Stepanov, a computer-vision and machine-learning engineer with experience in image processing, deep learning, medical imaging, augmented reality, and efficient inference.
Selected experience
Applied research and engineering across imaging, vision, and machine learning.
Publications
02
Computer vision for real-time systems
Development and deployment of detection, recognition, tracking, segmentation, and classification systems under practical accuracy, latency, and hardware constraints.
03
Imaging and measurement systems
Algorithm development, evaluation, and integration for optical measurement, medical imaging, camera-projector systems, and image-enhancement applications.
Working principles
Sound measurements. Reproducible methods. Engineering that works beyond the prototype.