InfoMatch: Entropy Neural Estimation for Semi-Supervised Image Classification

Authors: Qi Han, Zhibo Tian, Chengwei Xia, Kun Zhan

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

Reproducibility Variable Result LLM Response
Research Type Experimental Through extensive experiments, we show its superior performance. ... We evaluate Info Match on well-known benchmark datasets, including CIFAR-10/100 [Krizhevsky and Hinton, 2009], SVHN [Netzer et al., 2011], STL-10 [Coates et al., 2011], and Image Net [Deng et al., 2009].
Researcher Affiliation Academia Qi Han, Zhibo Tian, Chengwei Xia, Kun Zhan School of Information Science and Engineering, Lanzhou University kzhan@lzu.edu.cn
Pseudocode Yes Algorithm 1 The Info Match algorithm.
Open Source Code Yes The source code is available at https://github.com/kunzhan/Info Match.
Open Datasets Yes We evaluate Info Match on well-known benchmark datasets, including CIFAR-10/100 [Krizhevsky and Hinton, 2009], SVHN [Netzer et al., 2011], STL-10 [Coates et al., 2011], and Image Net [Deng et al., 2009].
Dataset Splits No The paper mentions 'Info Match performance is then evaluated using the EMA with a parameter of 0.999' and 'self-adaptive thresholding method' but does not specify explicit validation dataset splits (e.g., percentages or counts for a dedicated validation set).
Hardware Specification No The paper mentions using specific model architectures like Res Net-50 and Wide Res Net variants, but it does not provide any specific details about the hardware (e.g., GPU models, CPU types, or memory) used for running the experiments.
Software Dependencies No The paper does not provide specific software dependencies with version numbers (e.g., Python, PyTorch, CUDA versions).
Experiment Setup Yes Specifically, we employ standard stochastic gradient descent algorithm with cosine learning rate decay as the optimizer across all datasets, with an initial learning rate of 0.03 and a momentum of 0.9. For all experiments, we set the total number of iterations to 220. ... for Image Net, we maintain a batch size of 128 for both labeled and unlabeled samples, i.e., nb l = nb u = 128 ... For other datasets, we adjust the batch sizes to nb l = 64 and nb u = 448... Subsequently, we adjust the parameter λ that regulates the entropy bounds to 0.002.