Summary of Hierarchical State Space Models For Continuous Sequence-to-sequence Modeling, by Raunaq Bhirangi et al.
Hierarchical State Space Models for Continuous Sequence-to-Sequence Modeling
by Raunaq Bhirangi, Chenyu Wang, Venkatesh Pattabiraman, Carmel Majidi, Abhinav Gupta, Tess Hellebrekers, Lerrel Pinto
First submitted to arxiv on: 15 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Robotics (cs.RO); Signal Processing (eess.SP)
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 This paper presents Hierarchical State-Space Models (HiSS), a new technique for continuous sequential prediction. The authors address the challenges of using raw sensory data to predict desirable physical quantities, such as force or inertial measurements. They describe HiSS as stacking structured state-space models on top of each other to create a temporal hierarchy. The approach outperforms state-of-the-art sequence models, including causal Transformers, LSTMs, S4, and Mamba, by at least 23% on MSE across six real-world sensor datasets. The authors also demonstrate efficient scaling to smaller datasets and compatibility with existing data-filtering techniques. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about a new way to use raw sensory data to make predictions. Imagine you’re trying to predict what’s happening in the world just by looking at a bunch of numbers that come from sensors like magnetometers or piezoresistors. This is hard because these sensors can be noisy and affected by things outside what we want to predict. The authors came up with a new approach called Hierarchical State-Space Models (HiSS) that does better than other methods for this task. They tested it on six different datasets and showed that it’s at least 23% better than the existing best methods. This is important because making predictions from raw sensory data can help us with things like medical devices, robotics, and more. |
Keywords
* Artificial intelligence * Mse
