AURA: Visually Interpretable Affective Understanding via Robust Archetypes

Guanyu Hu, Dimitrios Kollias, Xinyu Yang
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:45181-45211, 2026.

Abstract

Interpretable methods such as vision–language models (VLMs) have advanced affect analysis by aligning images with textual descriptions. However, relying on text as an intermediate proxy faces critical limitations: linguistic templates are inherently discrete, making them fundamentally incompatible with continuous valence–arousal regression, while also acting as a bottleneck for fine-grained visual nuances. Cognitive psychology suggests that human affective perception is not mediated by linguistic translation, but is grounded in perceptual resemblance to internalized visual archetypes. Motivated by this, we propose AURA, an archetype-based framework that replaces brittle linguistic proxies with a self-organizing archetype manifold. By adaptively allocating representational density according to affective complexity, AURA enables accurate continuous regression and reshapes affective taxonomies by decomposing labels into interpretable, geometrically coherent visual primitives. This paradigm offers a transparent, visually grounded decision trail and achieves state-of-the-art results across discrete and continuous tasks.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-hu26z, title = {{AURA}: Visually Interpretable Affective Understanding via Robust Archetypes}, author = {Hu, Guanyu and Kollias, Dimitrios and Yang, Xinyu}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {45181--45211}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/hu26z/hu26z.pdf}, url = {https://proceedings.mlr.press/v306/hu26z.html}, abstract = {Interpretable methods such as vision–language models (VLMs) have advanced affect analysis by aligning images with textual descriptions. However, relying on text as an intermediate proxy faces critical limitations: linguistic templates are inherently discrete, making them fundamentally incompatible with continuous valence–arousal regression, while also acting as a bottleneck for fine-grained visual nuances. Cognitive psychology suggests that human affective perception is not mediated by linguistic translation, but is grounded in perceptual resemblance to internalized visual archetypes. Motivated by this, we propose AURA, an archetype-based framework that replaces brittle linguistic proxies with a self-organizing archetype manifold. By adaptively allocating representational density according to affective complexity, AURA enables accurate continuous regression and reshapes affective taxonomies by decomposing labels into interpretable, geometrically coherent visual primitives. This paradigm offers a transparent, visually grounded decision trail and achieves state-of-the-art results across discrete and continuous tasks.} }
Endnote
%0 Conference Paper %T AURA: Visually Interpretable Affective Understanding via Robust Archetypes %A Guanyu Hu %A Dimitrios Kollias %A Xinyu Yang %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-hu26z %I PMLR %P 45181--45211 %U https://proceedings.mlr.press/v306/hu26z.html %V 306 %X Interpretable methods such as vision–language models (VLMs) have advanced affect analysis by aligning images with textual descriptions. However, relying on text as an intermediate proxy faces critical limitations: linguistic templates are inherently discrete, making them fundamentally incompatible with continuous valence–arousal regression, while also acting as a bottleneck for fine-grained visual nuances. Cognitive psychology suggests that human affective perception is not mediated by linguistic translation, but is grounded in perceptual resemblance to internalized visual archetypes. Motivated by this, we propose AURA, an archetype-based framework that replaces brittle linguistic proxies with a self-organizing archetype manifold. By adaptively allocating representational density according to affective complexity, AURA enables accurate continuous regression and reshapes affective taxonomies by decomposing labels into interpretable, geometrically coherent visual primitives. This paradigm offers a transparent, visually grounded decision trail and achieves state-of-the-art results across discrete and continuous tasks.
APA
Hu, G., Kollias, D. & Yang, X.. (2026). AURA: Visually Interpretable Affective Understanding via Robust Archetypes. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:45181-45211 Available from https://proceedings.mlr.press/v306/hu26z.html.

Related Material