May 16, 2022 · In our PCL, we propose to generate the categorical classifiers based on the prototypes by performing a learnable mapping function. To further ...

Feb 1, 2023 · In this work, we show that learning prototype classifiers addresses the biased softmax problem in LTR. Prototype classifiers can deliver promising results.

May 16, 2022 · Abstract In this paper, we tackle the long-tailed visual recognition problem from the categorical prototype.

In this paper, we tackle the long-tailed visual recognition problem from the categorical prototype perspective by proposing a prototype-based classifier ...

(1) We propose a novel learnable prototype classifier for long-tailed recognition to counter the shortcoming of softmax to correlate classifier norm to class ...

This work shows that learning prototype classifiers addresses the biased softmax problem in LTR, and proposes to jointly learn prototypes by using distances ...

We propose a prototype-based contrastive learning(PCL) loss and prototype-based feature augmentation(PFA) module to improve the accuracy of the classifier on ...

Aug 19, 2023 · Further, we theoretically analyze the properties of Euclidean distance based prototype classifiers that lead to stable gradient-based ...

May 15, 2024 · To enable independent distance scales along each channel, we enhance Prototype classifiers by learning channel-dependent temperature parameters.

Jul 25, 2024 · Present long-tailed recognition methods usually use a step-by-step or progressive learning strategy to shift the training process from feature ...