Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].

OT-DETECTOR: Delving into Optimal Transport for Zero-shot Out-of-Distribution Detection

Authors: Yu Liu, Hao Tang, Haiqi Zhang, Jing Qin, Zechao Li

IJCAI 2025 | Venue PDF | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments on several benchmarks demonstrate that OT-DETECTOR achieves state-of-the-art performance across various OOD detection tasks, particularly in challenging hard-OOD scenarios.
Researcher Affiliation Academia Yu Liu1 , Hao Tang2 , Haiqi Zhang1 , Jing Qin2 and Zechao Li1, 1School of Computer Science and Engineering, Nanjing University of Science and Technology 2Centre for Smart Health, The Hong Kong Polytechnic University EMAIL, EMAIL
Pseudocode Yes Algorithm 1: Semantic-aware Content Refinement Input : Input image x, ID labels Yid, Image encoder I, Text encoder T , Confidence function M, Top-k selection parameter k, Number of views N Output: Refined feature f r
Open Source Code No The paper does not contain any explicit statement about providing open-source code for the described methodology, nor does it provide a link to a code repository.
Open Datasets Yes Following the previous work [Ming et al., 2022], we evaluate our method on the Image Net-1K OOD benchmark to ensure comparability with other methods. The Image Net-1K OOD benchmark uses Image Net-1K as the ID data and considers Texture [Cimpoi et al., 2014], i Naturalist [Horn et al., 2018], SUN [Xiao et al., 2010], and Places365 [Zhou et al., 2018] as OOD data.
Dataset Splits Yes Following the previous work [Ming et al., 2022], we evaluate our method on the Image Net-1K OOD benchmark to ensure comparability with other methods. The Image Net-1K OOD benchmark uses Image Net-1K as the ID data and considers Texture [Cimpoi et al., 2014], i Naturalist [Horn et al., 2018], SUN [Xiao et al., 2010], and Places365 [Zhou et al., 2018] as OOD data. Additionally, we perform hard OOD analysis using semantically similar subsets of Image Net constructed by MCM, i.e., Image Net-10, Image Net-20, and Image Net-100, which show high semantic similarity.
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, memory amounts) used for running its experiments.
Software Dependencies No The paper mentions using 'CLIP' and 'CLIP-B/16' as the pretrained model and backbone, and refers to 'Open AI s open-source models', but does not provide specific version numbers for any software dependencies like Python, deep learning frameworks, or libraries.
Experiment Setup Yes For view augmentation, N = 256 randomly cropped images are generated, and the Top-k = 20 crop features are selected for fusion. In the OT component, we fix ϵ = 90 for entropic regularization and dynamically determine the optimal α for each OOD dataset.