# Machine learning for N-dimensional spatial reasoning tasks on the web

## TL;DR

This paper introduces Snake-ML, a web-based tool for training spatial reasoning tasks using machine learning, which is efficient and adaptable for various applications.

## Contribution

The novel contribution is the development of Snake-ML, a client-side framework for efficient spatial reasoning training with significant speed improvements.

## Key findings

- Snake-ML achieves a 4.58× speedup in model inference on the edge.
- A TensorFlow.js GPU pipeline provides up to a 32× speedup in training time.
- The framework is adaptable to complex spatial tasks like autonomous systems and robotics.

## Abstract

Spatial reasoning is essential for solving complex tasks in dynamic and high-dimensional environments. However, current training models for spatial tasks are computationally demanding and heavily reliant on human input. To address this gap, we present Snake-ML, a web-based simulation tool and proof-of-concept framework designed to demonstrate client-side training of spatial reasoning tasks. Snake-ML serves as an efficient and intuitive test bed for developing spatial navigation strategies in browser-based environments. We chose the snake game as our test bed because it is well suited for demonstrating spatial reasoning in low-dimensional visual spaces while remaining relevant to higher-dimensional tasks, compared to alternative methods. Through quantitative analysis, on the edge alone, Snake-ML achieves a 4.58× speedup in model inference. Additionally, we developed a direct TensorFlow.js GPU pipeline that achieves up to a 32× speedup in training time without any CPU/GPU synchronization. This pipeline has the potential to improve many edge-based AI visualization projects. Snake-ML shows potential for adaptability to complex spatial tasks, such as autonomous systems, robotics, and AI-driven environments. Our code and web-based simulation tool are publicly available.