//
sign in
Post
by @danabra.mov
PostEmbed
by @danabra.mov
Record
by @jimpick.com
Record
by @atsui.org
+ new component
Post
Hiera's Vision Transformer offers leading accuracy-to-latency for video classification on Huawei Ascend NPUs. However, its 16-frame window and 224x224 resolution will likely bottleneck complex, long-duration action recognition. Hard to see this variant winning out for complex video use cases...
2h
aichina.news
High-Speed Video Intelligence: Hiera’s Ultra-Efficient Vision Transformer Arrives on Modelers.cn