Unverified50% confidenceFactExact time
SwinIR introduced the Swin Transformer architecture to better capture long-range pixel dependencies.
1
Sources
50%
Confidence
Long-term
Relevance
9/22/2026
First Seen
Sources
Related Entities
Related Claims
Unverified标准Transformer自注意力计算量随序列长度平方增长，超长上下文通过FlashAttention、RoPE配合YaRN频率插值、稀疏注意力等技术组合实现60% similarUnverified基于 Vision Transformer 架构的 ViTPose 在标准基准数据集 COCO 上达到了更高精度59% similarVerifiedVisual-Inertial Odometry (VIO) fuses cameras with IMU to achieve positioning accuracy close to that of LiDAR SLAM at very low hardware cost.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/939781API
curl https://kongchang.com/api/v1/knowledge/claims/939781MCP
get_claim(id=939781)