Delve into Base-Novel Confusion: Redundancy Exploration for Few-Shot Class-Incremental Learning

Authors: Haichen Zhou, Yixiong Zou, Ruixuan Li, Yuhua Li, Kui Xiao

IJCAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments across benchmark datasets, including CIFAR-100, mini Image Net, and CUB-2002011 demonstrate that our method achieves stateof-the-art performance.
Researcher Affiliation Academia Haichen Zhou1 , Yixiong Zou1 , Ruixuan Li1 , Yuhua Li1 and Kui Xiao2 1Huazhong University of Science and Technology 2Hubei University
Pseudocode No The paper describes its method in prose and mathematical equations but does not include any explicitly labeled 'Pseudocode' or 'Algorithm' blocks.
Open Source Code No The paper does not contain any explicit statements about releasing source code or provide links to a code repository.
Open Datasets Yes Our method is evaluated on CIFAR-100 [Krizhevsky et al., 2009], mini Image Net [Russakovsky et al., 2015], and CUB200 [Wah et al., 2011].
Dataset Splits No The paper describes the incremental learning setup (e.g., 60 base classes and 40 novel classes across 8 incremental sessions), but it does not specify explicit train/validation/test dataset splits in terms of percentages or fixed sample counts that are commonly provided for reproducibility.
Hardware Specification No The paper does not provide specific details about the hardware used for running experiments, such as GPU models, CPU specifications, or memory.
Software Dependencies No The paper mentions utilizing ResNet-12 and pretrained ResNet-18 models but does not specify the versions of any software libraries, frameworks (e.g., PyTorch, TensorFlow), or programming languages used.
Experiment Setup Yes Our training protocol follows [Yang et al., 2023], utilizing the Res Net-12 for CIFAR-100 and mini Image Net, and pretrained Res Net-18 for CUB200. Batch sizes of 128 and 100 are employed for the base session and incremental sessions, respectively. ... λ and β are hyper-parameres.