TensorFlores: An enhanced Python-based TinyML framework

• Tipo: Artigo em Revista Científica Internacional
• Data: 2025

Autores

Flores, TKS (Autor) Outra; Costa, DG (Autor) FEUP; Silva, I (Autor) Outra

Revista

Título: SoftwareX

Vol. 31

Editora: Elsevier

Indexação

ISI Web of Knowledge - 0 Citações

Scopus - 0 Citações

Outras Informações

ID Authenticus: P-018-YK5

DOI: 10.1016/j.softx.2025.102224

Abstract (EN): The TensorFlores framework is a Python-based tool designed to optimize machine learning deployment in resource-constrained environments. It introduces evolving clustering-based quantization, supporting both quantization-aware training and post-training quantization while maintaining model accuracy. TensorFlores converts TensorFlow MLP models into optimized formats and generates platform-agnostic C++ code for embedded systems. Its modular architecture minimizes memory usage and computational overhead, enabling efficient real-time inference. By combining clustering-based quantization and automated code generation, TensorFlores enhances TinyML applications, making it a robust solution for low-power and edge AI scenarios in embedded and IoT systems.

Idioma: Inglês

Tipo (Avaliação Docente): Científica

Nº de páginas: 8