Fully-Binarized Distance Computation Based On-Device Few-Shot Learning for XR Applications

Vivek Parmar, Sandeep Kaur Kingra, Syed Shakib Sarwar, Ziyun Li, Barbara De Salvo, Manan Suri

CVPRW 2023 pp. 4502-4508

doi:10.1109/CVPRW59228.2023.00473 /cvprw/2023/parmar2023cvprw-fullybinarized/

Abstract

Low-Power Edge-AI capabilities are essential for on-device extended reality (XR) applications to support the vision of Metaverse. A critical requirement for emerging AI applications is personalization and adaptability without requiring retraining. Few-shot learning using embedding-based computations present an attractive method for the same. However, quantization-based optimizations to map such computations are yet to be explored. In this work, we present a fully binarized distance computing (BinDC) framework to perform distance computations for few-shot learning using only accumulation and logic operations (XOR/XNOR). The proposed method leads to marginal loss in accuracy of ≈ 4% (for 4-bits). This leads to savings in memory (≈ 8 ), energy (≈ 2.5-3×), power (≈ 2×) and latency (≈ 1.1-1.5×) compared to a floating-point cosine distance computation when using CPU-based computations performed on an embedded platform. We further demonstrate realization utilizing RRAM (resistive random access memory) based IMC (in-memory computing) to further improve EDP (energy delay product) (≈ 1000×) in comparison to the embedded CPU-based realization.

CVPRW Semantic Scholar

Cite

Text

Parmar et al. "Fully-Binarized Distance Computation Based On-Device Few-Shot Learning for XR Applications." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2023. doi:10.1109/CVPRW59228.2023.00473

Markdown

[Parmar et al. "Fully-Binarized Distance Computation Based On-Device Few-Shot Learning for XR Applications." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2023.](https://mlanthology.org/cvprw/2023/parmar2023cvprw-fullybinarized/) doi:10.1109/CVPRW59228.2023.00473

BibTeX

@inproceedings{parmar2023cvprw-fullybinarized,
  title     = {{Fully-Binarized Distance Computation Based On-Device Few-Shot Learning for XR Applications}},
  author    = {Parmar, Vivek and Kingra, Sandeep Kaur and Sarwar, Syed Shakib and Li, Ziyun and De Salvo, Barbara and Suri, Manan},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
  year      = {2023},
  pages     = {4502-4508},
  doi       = {10.1109/CVPRW59228.2023.00473},
  url       = {https://mlanthology.org/cvprw/2023/parmar2023cvprw-fullybinarized/}
}