Ting-Wei Yu

Papers

Total Citations

H-Index

About

Ting-Wei Yu is a leading researcher in computer vision and graphics, specializing in photorealistic indoor scene reconstruction and dataset generation. His most impactful contribution is the **OpenRooms framework**, a groundbreaking open-source pipeline that transforms raw 3D scans into large-scale, photorealistic indoor datasets with precise ground truth for geometry, materials, lighting, and semantics. This work, published in 2021 and cited over 66 times, addresses a critical bottleneck in scene understanding by making high-quality, physically accurate data accessible to the broader research community. An earlier version of the framework (2020) has also garnered 13 citations, underscoring its sustained influence. By democratizing dataset creation, Yu’s research enables advances in inverse rendering, relighting, and embodied AI, where realistic training data is essential. His work is notable for its end-to-end approach—from raw scan to fully annotated, photorealistic scene—bridging the gap between synthetic and real-world data. For students and researchers, Yu’s contributions offer a powerful tool for training robust models in indoor scene understanding, lighting estimation, and material recognition, making him a key figure in the push toward more accessible, high-fidelity computer vision benchmarks.