PhD Candidate · Computer Vision · UFPR
I am a PhD candidate in Computer Science at the Federal University of Paraná (UFPR), where I work on computer vision, image super-resolution, and visual recognition. My research investigates how recognition systems can remain useful when image quality is limited by distance, motion, compression, weather, acquisition hardware, or other real-world conditions. I develop super-resolution models, task-aware objectives, datasets, and evaluation protocols that connect image restoration with downstream recognition, particularly for automatic license plate recognition and intelligent transportation systems. I am also interested in reproducible research through public benchmarks, transparent evaluation, and open-source implementations.
01 · Selected publications
Featured work
Introduces UFPR-SR-Plates, a benchmark with 10,000 real-world tracks and 100,000 paired low- and high-resolution license plate images.
Presents layout- and character-aware learning strategies for reconstructing license plates while preserving recognition-relevant information.
Combines attention and sub-pixel operations with an OCR-informed objective for heavily degraded license plate images.
Studies attention-based single-image super-resolution under severe degradation and releases the RodoSol-LR-HR dataset.
02 · Open research
Dataset · Benchmark
Paired real-world license plate imagery for super-resolution, temporal analysis, and recognition research.
Repository ↗Model · Source code
Layout-aware and character-driven license plate super-resolution with OCR-guided learning.
Repository ↗Dataset
Controlled low- and high-resolution image pairs for evaluating reconstruction across degradation levels.
Repository ↗03 · Contact