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/

Global: Zendy Plus annual

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

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Compared to the great progress of large-scale vision transformers (ViTs) inrecent years, large-scale models based on convolutional neural networks (CNNs)are still in an early state. This work presents a new large-scale CNN-basedfoundation model, termed InternImage, which can obtain the gain from increasingparameters and training data like ViTs. Different from the recent CNNs thatfocus on large dense kernels, InternImage takes deformable convolution as thecore operator, so that our model not only has the large effective receptivefield required for downstream tasks such as detection and segmentation, butalso has the adaptive spatial aggregation conditioned by input and taskinformation. As a result, the proposed InternImage reduces the strict inductivebias of traditional CNNs and makes it possible to learn stronger and morerobust patterns with large-scale parameters from massive data like ViTs. Theeffectiveness of our model is proven on challenging benchmarks includingImageNet, COCO, and ADE20K. It is worth mentioning that InternImage-H achieveda new record 65.4 mAP on COCO test-dev and 62.9 mIoU on ADE20K, outperformingcurrent leading CNNs and ViTs. The code will be released athttps://github.com/OpenGVLab/InternImage.

InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutions