@misc{indiciae2a2e13cea5b7,
  title = {Enabling Memory-efficient Im2win Convolution with Multi-precision Support on GPU CUDA and Tensor Cores},
  author = {Xiang Fu and Jixiang Ma and Xinpeng Zhang and Peng Zhao and Shuai Lu and Xu Tony Liu},
  year = {2026},
  url = {https://arxiv.org/abs/2608.20725},
  note = {Source identifier: 2608.20725}
}