Audio-Visual Generalized Zero-Shot Learning using Pre-Trained Large Multi-Modal Models
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916198473531392 |
|---|---|
| author | Kurzendörfer, David Mercea, Otniel-Bogdan Koepke, A. Sophia Akata, Zeynep |
| author_facet | Kurzendörfer, David Mercea, Otniel-Bogdan Koepke, A. Sophia Akata, Zeynep |
| contents | Audio-visual zero-shot learning methods commonly build on features extracted from pre-trained models, e.g. video or audio classification models. However, existing benchmarks predate the popularization of large multi-modal models, such as CLIP and CLAP. In this work, we explore such large pre-trained models to obtain features, i.e. CLIP for visual features, and CLAP for audio features. Furthermore, the CLIP and CLAP text encoders provide class label embeddings which are combined to boost the performance of the system. We propose a simple yet effective model that only relies on feed-forward neural networks, exploiting the strong generalization capabilities of the new audio, visual and textual features. Our framework achieves state-of-the-art performance on VGGSound-GZSL, UCF-GZSL, and ActivityNet-GZSL with our new features. Code and data available at: https://github.com/dkurzend/ClipClap-GZSL. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_06309 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Audio-Visual Generalized Zero-Shot Learning using Pre-Trained Large Multi-Modal Models Kurzendörfer, David Mercea, Otniel-Bogdan Koepke, A. Sophia Akata, Zeynep Computer Vision and Pattern Recognition Audio-visual zero-shot learning methods commonly build on features extracted from pre-trained models, e.g. video or audio classification models. However, existing benchmarks predate the popularization of large multi-modal models, such as CLIP and CLAP. In this work, we explore such large pre-trained models to obtain features, i.e. CLIP for visual features, and CLAP for audio features. Furthermore, the CLIP and CLAP text encoders provide class label embeddings which are combined to boost the performance of the system. We propose a simple yet effective model that only relies on feed-forward neural networks, exploiting the strong generalization capabilities of the new audio, visual and textual features. Our framework achieves state-of-the-art performance on VGGSound-GZSL, UCF-GZSL, and ActivityNet-GZSL with our new features. Code and data available at: https://github.com/dkurzend/ClipClap-GZSL. |
| title | Audio-Visual Generalized Zero-Shot Learning using Pre-Trained Large Multi-Modal Models |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2404.06309 |