ExpPoint-MAE: Better interpretability and performance for self-supervised point cloud transformers
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2023
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866914751891636224 |
|---|---|
| author | Romanelis, Ioannis Fotis, Vlassis Moustakas, Konstantinos Munteanu, Adrian |
| author_facet | Romanelis, Ioannis Fotis, Vlassis Moustakas, Konstantinos Munteanu, Adrian |
| contents | In this paper we delve into the properties of transformers, attained through self-supervision, in the point cloud domain. Specifically, we evaluate the effectiveness of Masked Autoencoding as a pretraining scheme, and explore Momentum Contrast as an alternative. In our study we investigate the impact of data quantity on the learned features, and uncover similarities in the transformer's behavior across domains. Through comprehensive visualiations, we observe that the transformer learns to attend to semantically meaningful regions, indicating that pretraining leads to a better understanding of the underlying geometry. Moreover, we examine the finetuning process and its effect on the learned representations. Based on that, we devise an unfreezing strategy which consistently outperforms our baseline without introducing any other modifications to the model or the training pipeline, and achieve state-of-the-art results in the classification task among transformer models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2306_10798 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | ExpPoint-MAE: Better interpretability and performance for self-supervised point cloud transformers Romanelis, Ioannis Fotis, Vlassis Moustakas, Konstantinos Munteanu, Adrian Computer Vision and Pattern Recognition Machine Learning In this paper we delve into the properties of transformers, attained through self-supervision, in the point cloud domain. Specifically, we evaluate the effectiveness of Masked Autoencoding as a pretraining scheme, and explore Momentum Contrast as an alternative. In our study we investigate the impact of data quantity on the learned features, and uncover similarities in the transformer's behavior across domains. Through comprehensive visualiations, we observe that the transformer learns to attend to semantically meaningful regions, indicating that pretraining leads to a better understanding of the underlying geometry. Moreover, we examine the finetuning process and its effect on the learned representations. Based on that, we devise an unfreezing strategy which consistently outperforms our baseline without introducing any other modifications to the model or the training pipeline, and achieve state-of-the-art results in the classification task among transformer models. |
| title | ExpPoint-MAE: Better interpretability and performance for self-supervised point cloud transformers |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2306.10798 |