English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

3D point-clouds and 2D images are different visual representations of thephysical world. While human vision can understand both representations,computer vision models designed for 2D image and 3D point-cloud understandingare quite different. Our paper explores the potential of transferring 2D modelarchitectures and weights to understand 3D point-clouds, by empiricallyinvestigating the feasibility of the transfer, the benefits of the transfer,and shedding light on why the transfer works. We discover that we can indeeduse the same architecture and pretrained weights of a neural net model tounderstand both images and point-clouds. Specifically, we transfer theimage-pretrained model to a point-cloud model by copying or inflating theweights. We find that finetuning the transformed image-pretrained models (FIP)with minimal efforts -- only on input, output, and normalization layers -- canachieve competitive performance on 3D point-cloud classification, beating awide range of point-cloud models that adopt task-specific architectures and usea variety of tricks. When finetuning the whole model, the performance improveseven further. Meanwhile, FIP improves data efficiency, reaching up to 10.0top-1 accuracy percent on few-shot classification. It also speeds up thetraining of point-cloud models by up to 11.1x for a target accuracy (e.g., 90 %accuracy). Lastly, we provide an explanation of the image to point-cloudtransfer from the aspect of neural collapse. The code is available at:\url{https://github.com/chenfengxu714/image2point}.

Image2Point: 3D Point-Cloud Understanding with 2D Image Pretrained Models