Summary of A Trainable Feature Extractor Module For Deep Neural Networks and Scanpath Classification, by Wolfgang Fuhl
A Trainable Feature Extractor Module for Deep Neural Networks and Scanpath Classification
by Wolfgang Fuhl
First submitted to arxiv on: 19 Mar 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 paper proposes a trainable feature extraction module for deep neural networks in the context of scanpath classification, with potential applications in medicine, manufacturing, and training systems. The module transforms a scanpath into a feature vector usable by the neural network architecture. It adapts parameters based on backpropagated error to improve classification performance, making it jointly trainable with the neural network. The module is motivated by classical histogram-based approaches that compute distributions over a scanpath. The authors evaluate their approach on three public datasets and compare it to state-of-the-art methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper creates a feature extraction module for deep neural networks in eye tracking research, which can be used in medicine, manufacturing, and education. This module takes a scanpath (a sequence of eye movements) and turns it into a format that the neural network can understand. The module learns to improve its performance by using errors from the neural network. This approach is inspired by older methods that calculate distributions over a scanpath. The authors test their method on three public datasets and compare it to other similar approaches. |
Keywords
* Artificial intelligence * Classification * Feature extraction * Neural network * Tracking
