.png){kind=icon}
.png){kind=icon}
Return
Dynamic collaborative learning with heterogeneous knowledge transfer for long-tailed visual recognition
.png){kind=icon}
PRE
AI
DOI：10.1016/j.inffus.2024.102734.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Solving the long-tailed visual recognition with deep convolutional neural networks is still a challenging task. As a mainstream method, multi-experts models achieve SOTA accuracy for tackling this problem, the uncertainty in network learning and the complexity infusion inference constrain the performance practicality of the multi-experts models. To remedy this, we propose a novel dynamic collaborative learning with heterogeneous knowledge transfer model (DCHKT) in this paper, in which experts with different expertise collaborate to make predictions. DCHKT consists of two core components: dynamic adaptive weight adjustment and heterogeneous knowledge transfer learning. First, the dynamic adaptive weight adjustment is designed to shift the focus of model training between the global expert and domain experts via dynamic adaptive weight. By modulating the trade-off between the learning of features and classifier, the dynamic adaptive weight adjustment can enhance the discriminative ability of each expert and alleviate the uncertainty model learning. Then, heterogeneous knowledge transfer learning, which measures the distribution differences between the fusion logits of multiple experts and the predicted logits of each expert with different specialties, can achieve message passing between experts and enhance the consistency of ensemble prediction in model training and inference to promote their collaborations. Finally, extensive experimental results on public longtailed datasets: CIFAR-LT, ImageNet-LT, Place-LT and iNaturalist2018, demonstrate the effectiveness superiority of our DCHKT.
Keywords:
Long-tailed recognition
Heterogeneous knowledge transfer
Dynamic adaptive sinusoidal weight
Multi-experts collaboration
Inference uncertainty and complexity
Journal
IF:
15.5
Papers:
4.1K
Citations:
2.7W