Summary of Zero-shot Classification Using Hyperdimensional Computing, by Samuele Ruffino et al.
Zero-shot Classification using Hyperdimensional Computing
by Samuele Ruffino, Geethan Karunaratne, Michael Hersche, Luca Benini, Abu Sebastian, Abbas Rahimi
First submitted to arxiv on: 30 Jan 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed Hyperdimensional Computing Zero-shot Classifier (HDC-ZSC) is an innovative approach for zero-shot learning-based classification. This model combines a trainable image encoder with an attribute encoder inspired by Hyperdimensional Computing (HDC). The HDC-ZSC consists of three components: a trainable image encoder, an attribute encoder based on HDC, and a similarity kernel. It achieves Pareto optimal results on the CUB-200 dataset with a top-1 classification accuracy of 63.8% using only 26.6 million parameters. This outperforms two state-of-the-art non-generative approaches while requiring fewer parameters. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The Hyperdimensional Computing Zero-shot Classifier is a new way for computers to learn and classify things without seeing examples before. It uses special codes to help the computer understand what’s in a picture, even if it’s never seen that type of picture before. This model does two tasks: first, it identifies the important features of an image, and then it classifies the image into a category. The results are impressive – it can correctly classify 63.8% of images without any prior training. |
Keywords
* Artificial intelligence * Classification * Encoder * Zero shot
