Introduction

[中文]

ESP-DL is a library for high-performance deep learning resources dedicated to ESP32, ESP32-S2, ESP32-S3 and ESP32-C3.

Overview

ESP-DL provides APIs for Neural Network (NN) Inference, Image Processing, Math Operations and some Deep Learning Models. With ESP-DL, you can use Espressif’s SoCs for AI applications easily and fast.

As ESP-DL does not need any peripherals, it can be used as a component of some projects. For example, you can use it as a component of ESP-WHO, which contains several project-level examples of image application. The figure below shows what ESP-DL consists of and how ESP-DL is implemented as a component in a project.

Architecture Overview

Get Started with ESP-DL

For setup instructions to get started with ESP-DL, please read Getting Started with ESP-DL.

Please use the ESP-IDF on version 5.0 or above.

Try Models in the Model Zoo

ESP-DL provides some model APIs in the model_zoo , such as Human Face Detection, Human Face Recognition, Cat Face Detection, etc. You can use these models in the table below out of box.

Name	API Example
Human Face Detection	human_face_detect
Human Face Recognition	face_recognition
Cat Face Detection	cat_face_detect

Deploy a Model

To deploy a model, please proceed to Tutorials, where the instructions with three runnable examples will quickly help you design your model.

When you read the instructions, the following materials might be helpful:

• DL API

  • Variables and Constants: information about

    • variable: tensors

    • constants: filters, biases, and activations

  • Customizing Layers: instructions on how to customize a layer.

  • API Documentation : guides to provided API about Layer, Neural Network (NN), Math and tools.

    For API documentation, please refer to annotations in header files for the moment.

• Platform Conversion

  • TVM(Recommended)： Use AI compiler TVM to deploy AI model，More information about TVM please refer to TVM

  • Quantization Toolkit: a tool for quantizing floating-point models and evaluating quantized models on ESP SoCs

    • Toolkit: see Quantization Toolkit Overview

    • Toolkit API: see Quantization Toolkit APIs

  • Convert Tool: the tool and configuration file for floating-point quantization on coefficient.npy

    • config.json: see Specification of config.json

    • convert.py: see Usage of convert.py

      convert.py requires Python 3.7 or versions higher.

• Software and Hardware Boost

  • Quantization Specification: rules of floating-point quantization

Feedback

For feature requests or bug reports, please submit an issue. We will prioritize the most anticipated features.