A Modular 3D Face Reconstruction Benchmark: M3DFB paper presented in IEEE FG’25
Paper (arXiv) 📄 / Project Page 🌐 / Poster 📌 / Code 💻
Facial Basis paper (Beyond FACS) accepted and presented in IEEE FG’25
Project Page 🌐 / Poster 📌 / Code 💻
Our 3D reconstruction method (3DI) was published in TPAMI
3DI was published on Feb’24 issue of T-PAMI: https://ieeexplore.ieee.org/abstract/document/10330115 This paper shows how to implement the 3DI method efficiently, and provides experiments that show the ability of 3DI to capture person-specific facial morphology. Further, experiments show that 3DI stands out with its reconstruction performance that becomes increasingly more accurate as more frames per person are provided, whereas the performance of the compared deep learning-based methods tend to saturate earlier, with fewer number of frames.
Our paper on limitations of 3D reconstruction *benchmark metrics* was presented at IJCB’23
Our paper titled Meta-evaluation for 3D Face Reconstruction Via Synthetic Data was presented in an oral and also a poster presentation session of IJCB: https://ijcb2023.ieee-biometrics.org/oral-presentations/ Here is my blog post with the paper link and a summary of the study.