MPBR for Open-Set Fine-Grained Recognition
Jan 1, 2025
·
1 min read
MPBR explores recognition in open-world settings where unknown fine-grained categories appear at inference time. The method combines forward refinement of visual features with backward multimodal consistency constraints.
My work included developing benchmark pipelines and evaluating leading open-set-recognition models on iNat2021-OSR. Analysis used t-SNE, Grad-CAM, Swin Transformer backbones, and HAN-OSFGR.

Authors
AI for Science · Computational Chemistry · Computer Vision
Yu Zhu works at the intersection of AI for Science, computational chemistry,
and computer vision. His research spans machine learning force fields,
physics-driven reconstruction of molecular orbitals from scanning tunneling
microscopy images, open-set fine-grained recognition, and materials modeling.