BIDL User Guide

Contents

• Overview
  • Main Features
  • Software Architecture
  • Applicable Products
  • Basic Concepts
  • Restrictions
• Software Installation
  • Acquiring Software
  • Validation
  • Directory Description
• Quick Start
  • Choose Configuration File
  • Prepare Dataset
  • Model Training
  • Model Inference (GPU)
  • Model Inference (Lynxi brain-inspired systems)
• Introduction to Principles
  • ANN+SNN Hybrid Neural Network
  • Code Expression Form of Spatiotemporal Hybrid Network
  • Network Structures Supported by BIDL
  • BPTT Training of Hybrid Networks
  • Description of Supported Neuron Models in BIDL
• Designing flow
  • Training Part
  • Evaluation and Deployment to Lynxi Brain-inspired Chips (APU)
• API of Neuronal Models
  • Neuronal Models
• Datasets Introduction
  • Supported Datasets
• Case Study
  • Classification
  • Brain-inspired Tracking (Continuous Attractor)
  • DVS High-speed Target Detection
  • Location Recognition
  • Small Target Segmentation and Detection with PCNN
  • Frequency/Time/Population Coding
  • Inner Loop SNN
  • Summary of Used Networks in Above Cases
• Neuron Simulation Toolkit
  • Limitations
  • Installation Directory
  • Operation Instructions
  • Typical Cases
• Demonstration Scripts
  • Model Training (GPU)
  • Model Inference (GPU)
  • Model Inference (Lynxi chip)
  • Single Machine Multi-card Inference (Lynxi chip)
  • Model Multi-batch Inference (Lynxi chip)
  • Location Recognition
  • ST-YOLO Pedestrian Vehicle Detection (Lynxi chip)
  • DVS High-speed Rotary Target Detection (Lynxi chip)
  • Neural simulations
• Professional Version Introduction
• Contact Us