BROWSE
Detail
Conference Paper
EQ-CBM: A Probabilistic Concept Bottleneck with Energy-based Models and Quantized Vectors
Cited 0 time in 
Share 
Share
- Authors
- Sangwon Kim, Dasom Ahn, Byoung Chul Ko, In-su Jang, Kwang-Ju Kim
- Issue Date
- 2024-12
- Citation
- Asian Conference on Computer Vision (ACCV) 2024, pp.3432-3448
- Language
- English
- Type
- Conference Paper
- Abstract
- The demand for reliable AI systems has intensified the need for interpretable deep neural networks. Concept bottleneck models (CBMs) have gained attention as an effective approach by leveraging human-understandable concepts to enhance interpretability. However, existing CBMs face challenges due to deterministic concept encoding and reliance on inconsistent concepts, leading to inaccuracies. We propose EQ-CBM, a novel framework that enhances CBMs through probabilistic concept encoding using energy-based models (EBMs) with quantized concept activation vectors (qCAVs). EQ-CBM effectively captures uncertainties, thereby improving prediction reliability and accuracy. By employing qCAVs, our method selects homogeneous vectors during concept encoding, enabling more decisive task performance and facilitating higher levels of human intervention. Empirical results using benchmark datasets demonstrate that our approach outperforms the state-of-the-art in both concept and task accuracy.
- KSP Keywords
- Benchmark datasets, Deep neural network(DNN), Energy-based models, Prediction reliability, human intervention, need for, neural network(NN), state-of-The-Art, task performance