espressif/esp-dl

3.3.1

Latest
uploaded 3 hours ago
esp-dl is a lightweight and efficient neural network inference framework designed specifically for ESP series chips.

19 examples

  • how_to_deploy_streaming_model 1
  • cat_detect
    A simple image inference example using `cat.jpg` for testing detection results on ESP32-P4.
    226.36 KB
  • color_detect
    A simple image inference example.
    13.23 KB
  • dog_detect
    A simple image inference example using dog detection with specified configurations and output results.
    633.70 KB
  • hand_detect
    A simple image inference example using hand detection on images, demonstrating results before and after quantization.
    365.57 KB
  • hand_gesture_recognition
    A simple image inference example for recognizing 10 hand gestures plus a 'no_hand' category.
    750.97 KB
  • how_to_run_model
    12.81 KB
  • human_face_detect
    A simple image inference example demonstrating human face detection using ESP-IDF.
    42.52 KB
  • human_face_recognition
    A simple image inference example for human face recognition using ESP32-S3 and ESP32-P4 with configurable options.
    102.67 KB
  • mobilenetv2_cls
    A simple image inference example for ImageNet classification, demonstrating output from a quick start setup.
    38.99 KB
  • model_in_sdcard
    10.68 KB
  • motion_detect
    52.15 KB
  • pedestrian_detect
    A simple image inference example for pedestrian detection utilizing ESP-IDF. Follow the quick start guide to flash the example.
    96.93 KB
  • speaker_verification
    A simple audio inference example for speaker verification using three audio samples to compare similarities.
    681.76 KB
  • yolo11_detect
    A simple image inference example using Yolo11, demonstrating detection results before and after quantization.
    670.34 KB
  • yolo11_pose
    A simple image inference example using Yolo11, demonstrating pose recognition with ESP32-P4.
    266.98 KB
  • yolo26_detect
    This example enables quantized YOLOv26n inference on Espressif SoCs with high performance and flexible model loading.
    3.21 MB