SONA: Learning Conditional, Unconditional, and Mismatching-Aware Discriminator
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866912630637068288 |
|---|---|
| author | Takida, Yuhta Hayakawa, Satoshi Shibuya, Takashi Imaizumi, Masaaki Murata, Naoki Nguyen, Bac Uesaka, Toshimitsu Lai, Chieh-Hsin Mitsufuji, Yuki |
| author_facet | Takida, Yuhta Hayakawa, Satoshi Shibuya, Takashi Imaizumi, Masaaki Murata, Naoki Nguyen, Bac Uesaka, Toshimitsu Lai, Chieh-Hsin Mitsufuji, Yuki |
| contents | Deep generative models have made significant advances in generating complex content, yet conditional generation remains a fundamental challenge. Existing conditional generative adversarial networks often struggle to balance the dual objectives of assessing authenticity and conditional alignment of input samples within their conditional discriminators. To address this, we propose a novel discriminator design that integrates three key capabilities: unconditional discrimination, matching-aware supervision to enhance alignment sensitivity, and adaptive weighting to dynamically balance all objectives. Specifically, we introduce Sum of Naturalness and Alignment (SONA), which employs separate projections for naturalness (authenticity) and alignment in the final layer with an inductive bias, supported by dedicated objective functions and an adaptive weighting mechanism. Extensive experiments on class-conditional generation tasks show that \ours achieves superior sample quality and conditional alignment compared to state-of-the-art methods. Furthermore, we demonstrate its effectiveness in text-to-image generation, confirming the versatility and robustness of our approach. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_04576 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | SONA: Learning Conditional, Unconditional, and Mismatching-Aware Discriminator Takida, Yuhta Hayakawa, Satoshi Shibuya, Takashi Imaizumi, Masaaki Murata, Naoki Nguyen, Bac Uesaka, Toshimitsu Lai, Chieh-Hsin Mitsufuji, Yuki Machine Learning Artificial Intelligence Computer Vision and Pattern Recognition Deep generative models have made significant advances in generating complex content, yet conditional generation remains a fundamental challenge. Existing conditional generative adversarial networks often struggle to balance the dual objectives of assessing authenticity and conditional alignment of input samples within their conditional discriminators. To address this, we propose a novel discriminator design that integrates three key capabilities: unconditional discrimination, matching-aware supervision to enhance alignment sensitivity, and adaptive weighting to dynamically balance all objectives. Specifically, we introduce Sum of Naturalness and Alignment (SONA), which employs separate projections for naturalness (authenticity) and alignment in the final layer with an inductive bias, supported by dedicated objective functions and an adaptive weighting mechanism. Extensive experiments on class-conditional generation tasks show that \ours achieves superior sample quality and conditional alignment compared to state-of-the-art methods. Furthermore, we demonstrate its effectiveness in text-to-image generation, confirming the versatility and robustness of our approach. |
| title | SONA: Learning Conditional, Unconditional, and Mismatching-Aware Discriminator |
| topic | Machine Learning Artificial Intelligence Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2510.04576 |