Summary of Improving Personalisation in Valence and Arousal Prediction Using Data Augmentation, by Munachiso Nwadike et al.
Improving Personalisation in Valence and Arousal Prediction using Data Augmentation
by Munachiso Nwadike, Jialin Li, Hanan Salam
First submitted to arxiv on: 13 Apr 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
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 an enhanced personalization strategy for emotion recognition and Human-Machine Interaction (HMI) called Distance Weighting Augmentation (DWA). This approach leverages data augmentation to develop tailored models for continuous valence and arousal prediction. The method employs a weighting-based augmentation technique that expands a target individual’s dataset, identifying similar samples at the segment-level using distance metrics. Experimental results on the MuSe-Personalisation 2023 Challenge dataset demonstrate significant improvements in performance, particularly for features with low baseline performance, without sacrificing performance on high-performing features. The proposed method achieves a maximum combined testing CCC of 0.78, outperforming the reported baseline score of 0.76 (reproduced at 0.72). Additionally, it achieves peak arousal and valence scores of 0.81 and 0.76, respectively, surpassing reproduced baseline scores of 0.76 and 0.67. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper develops a new way to recognize emotions in people using computers. The goal is to create personalized models that work better for each individual. The problem is that sometimes there isn’t enough data for certain people, so the researchers came up with a solution called Distance Weighting Augmentation (DWA). This method takes a small amount of data from one person and makes it more like other people’s data, so it can be used to train better models. The results show that this approach works really well, especially for people who don’t have much data. It’s an important step forward in making computers understand people’s emotions better. |
Keywords
* Artificial intelligence * Data augmentation
