Revisiting class-incremental learning with pre-trained models: Generalizability and adaptivity are all you need
Class-incremental learning (CIL) aims to adapt to emerging new classes without forgetting
old ones. Traditional CIL models are trained from scratch to continually acquire knowledge …
old ones. Traditional CIL models are trained from scratch to continually acquire knowledge …
Fantastic gains and where to find them: On the existence and prospect of general knowledge transfer between any pretrained model
Training deep networks requires various design decisions regarding for instance their
architecture, data augmentation, or optimization. In this work, we find these training …
architecture, data augmentation, or optimization. In this work, we find these training …
Dual-curriculum teacher for domain-inconsistent object detection in autonomous driving
Object detection for autonomous vehicles has received increasing attention in recent years,
where labeled data are often expensive while unlabeled data can be collected readily …
where labeled data are often expensive while unlabeled data can be collected readily …
Dual branch network towards accurate printed mathematical expression recognition
Over the past years, Printed Mathematical Expression Recognition (PMER) has progressed
rapidly. However, due to the insufficient context information captured by Convolutional …
rapidly. However, due to the insufficient context information captured by Convolutional …
Class-incremental learning for baseband modulation classification: A comparison
This paper presents a comprehensive study on the capabilities of class-incremental learning
in the context of baseband modulation classification. Despite the growing interest in …
in the context of baseband modulation classification. Despite the growing interest in …
ATMKD: adaptive temperature guided multi-teacher knowledge distillation
Y Lin, S Yin, Y Ding, X Liang - Multimedia Systems, 2024 - Springer
Abstract Knowledge distillation is a technique that aims to distill the knowledge from a large
well-trained teacher model to a lightweight student model. In recent years, multi-teacher …
well-trained teacher model to a lightweight student model. In recent years, multi-teacher …
Correlation Guided Multi-teacher Knowledge Distillation
L Shi, N Jiang, J Tang, X Huang - International Conference on Neural …, 2023 - Springer
Abstract Knowledge distillation is a model compression technique that transfers knowledge
from a redundant and strong network (teacher) to a lightweight network (student). Due to the …
from a redundant and strong network (teacher) to a lightweight network (student). Due to the …
Mutually Promoted Hierarchical Learning for Incremental Implicitly-Refined Classification
G Zhao, Y Hou, K Mu - 2023 International Joint Conference on …, 2023 - ieeexplore.ieee.org
Class incremental learning devotes to learning a classification model from incrementally
arriving training data. Existing methods tend to use a single-headed layout due to the lack of …
arriving training data. Existing methods tend to use a single-headed layout due to the lack of …