
The core message of this work is to address the projection bias problem in generalized zero-shot learning (GZSL) by introducing a parameterized Mahalanobis distance metric to improve classification performance on both seen and unseen classes.


coremsg

overcoming-projection-bias-in-generalized-zero-shot-learning-through-parameterized-mahalanobis-distance-learning


Overcoming Projection Bias in Generalized Zero-Shot Learning through Parameterized Mahalanobis Distance Learning


title_rewrite


A novel Part Prototype Network (PPN) that leverages pre-trained Vision-Language detectors like VINVL to efficiently obtain region-specific attribute representations for improved Generalized Zero-Shot Learning performance.


leveraging-pre-trained-vision-language-detectors-for-efficient-generalized-zero-shot-learning


Leveraging Pre-Trained Vision-Language Detectors for Efficient Generalized Zero-Shot Learning



A novel Dual Expert Distillation Network (DEDN) approach that effectively models the inherent asymmetry of attributes and leverages both region and channel information to enhance generalized zero-shot learning performance.


dual-expert-distillation-network-for-improved-generalized-zero-shot-learning


Dual Expert Distillation Network for Improved Generalized Zero-Shot Learning
