Running LSTM neural networks on an Imagination NNA
January 8, 2021
Speech recognition has become more relevant in recent years: it enables computers to translate spoken language into text. It can be found in different types of applications, such as translators or closed captioning. An example of this technology is Mozilla’s DeepSpeech, an open-source speech-to-text engine, which uses a model trained by machine learning techniques based on Baidu’s Deep Speech research paper. We are going to provide an overview of how we are running version 0.5.1 of this model, by accelerating a static LSTM network on the Imagination neural network accelerator (NNA), with the goal of creating a prototype of a voice assistant for an automotive use case.
To read the full article, click here
Related Semiconductor IP
Related Blogs
- Running OpenCode with llama.cpp and Qwen 3.5 on E-Series
- Why the PowerVR 2NX NNA is the future of neural net acceleration
- Self-Compressing Neural Networks
- ChiPy®: Bridge Neural Networks and C++ on Silicon — Full Inference Pipelines with Zero CPU Round-Trips
Latest Blogs
- Tape-Out Readiness Checklist: Engineering Decisions That Prevent Costly Respins
- How Cadence DSPs Put In-Cabin AI Audio On-Chip in SemiDrive's X10
- The Fastest Path to Scalable Photonic Systems: Using Proven IP for both PIC and EIC designs
- Building Trusted AI Agents from the Silicon Root of Trust
- Heterogeneous Computing in Space