CodeProject.AI Modules
Supporting CodeProject.AI Server 2.9.2.
Computer Audition
- Sound Classifier (TensorFlow)
The sound classifier uses Tensorflow with Python to classify sound files based on the UrbanSound8K dataset.
v1.4.0 All Platforms Python, TensorFlow
Project by Chris Maunder, based on Tensorflow-Audio-Classification.
Computer Vision
- License Plate Reader
Detects and readers single-line and multi-line license plates using YOLO object detection and the PaddleOCR toolkit
v3.3.1 All Platforms except Windows-arm64 Python, PaddlePaddle
By Mike Lud - License Plate Reader (RKNN)
Detects and readers single-line and multi-line licence plates. This module only works with Rockchip RK3588/RK3588S NPUs like the Orange Pi 5/5B/5 Plus or Radxa ROCK.
v1.5.0 Orange Pi, Radxa ROCK Python, FastDeploy
By Mike Lud - Object Detection (Coral)
The object detection module uses the Coral TPU to locate and classify the objects the models have been trained on.
v2.4.0 All Platforms Python, TensorFlow-Lite
Project by Chris Maunder, Seth Price, based on Coral.ai examples. - Object Detection (YOLOv5 .NET)
Provides Object Detection using YOLOv5 ONNX models with DirectML. This module is best for those on Windows and Linux without CUDA enabled GPUs
v1.12.0 All Platforms except Windows-arm64 C#, ONNX, DirectML, YOLO
Project by Matthew Dennis, based on yolov5-net. - Object Detection (YOLOv5 3.1)
Provides Object Detection using YOLOv5 3.1 targeting CUDA 10 or 11 for older GPUs.
v1.12.0 All Platforms except macOS Python, PyTorch, YOLO
Project by Chris Maunder, Matthew Dennis, based on Deepstack. - Object Detection (YOLOv5 6.2)
Provides Object Detection using YOLOv5 6.2 targeting CUDA 11.5+, PyTorch < 2.0 for newer GPUs.
v1.10.0 All Platforms except Raspberry Pi, Jetson
Project by Matthew Dennis, based on Ultralytics YOLOv5. - Object Detection (YOLOv5 RKNN)
Provides Object Detection using YOLOv5 RKNN models. This module only works with Rockchip RK3588/RK3588S NPUs like the Orange Pi 5/5B/5 Plus
v1.8.1 Orange Pi, Radxa ROCK Python, FastDeploy, YOLO
By Mike Lud - Object Detection (YOLOv8)
Provides Object Detection in Python>=3.8 using YOLOv8. Great for newer NVIDIA GPUs
v1.6.0 All Platforms Python, PyTorch, YOLO
Project by Chris Maunder, based on ultralytics. - Optical Character Recognition
Provides OCR support using the PaddleOCR toolkit
v2.2.1 All Platforms except Windows-arm64 Python, PaddlePaddle
By Mike Lud - Scene Classification
Classifies an image according to one of 365 pre-trained scenes
v1.8.0 All Platforms except Jetson Python, PyTorch
Project by Chris Maunder, Matthew Dennis, based on Deepstack.
Face Recognition
- Face Processing
A number of Face image APIs including detect, recognize, and compare.
v1.12.1 All Platforms except Jetson Python, PyTorch
Project by Chris Maunder, Matthew Dennis, based on Deepstack.
Generative AI
- LlamaChat
A Large Language Model based on the Machine Learning Compilation for LLMs
v1.7.1 All Platforms except Windows-arm64, Raspberry Pi, Orange Pi, Radxa ROCK, Jetson Python, Llama
Project by Chris Maunder, based on llama-cpp-python. - MultiModeLLM
A multi-modal Large Language Model
v1.1.0 Windows, Linux, macOS, macOS Python, Phi-3
Project by Chris Maunder, based on chat-with-phi-3-vision. - Text to Image
Generates an image from a text prompt.
v1.3.1 Windows, macOS, Linux Python, PyTorch, Stable Diffusion
Project by Matthew Dennis, based on Diffusers.
Image Processing
- Background Remover
Automatically removes the background from a picture
v1.11.0 All Platforms except Linux, Raspberry Pi, Orange Pi, Radxa ROCK, Jetson Python, ONNX
Project by Chris Maunder, based on rembg. - Cartoonizer
Convert a photo into an anime style cartoon
v1.7.0 All Platforms except Raspberry Pi, Orange Pi, Radxa ROCK, Jetson Python, PyTorch
Project by Chris Maunder, based on animegan2-pytorch. - Portrait Filter
Provides a depth-of-field (bokeh) effect on images. Great for selfies.
v2.2.0 Windows C#, ONNX, DirectML
Project by Matthew Dennis, based on C# PortraitModeFilter. - Super Resolution
Increases the resolution of an image using AI
v2.2.1 All Platforms Python, ONNX
Project by Chris Maunder, based on PyTorch.org example.
Natural Language
- Sentiment Analysis
Provides an analysis of the sentiment of a piece of text. Positive or negative?
v2.2.0 Windows, macOS C#, TensorFlow
Project by Matthew Dennis, based on .NET ML Samples. - Text Summary
Summarizes text content by selecting a number of sentences that are most representative of the content.
v1.9.0 All Platforms Python, NLTK
Project by Chris Maunder, based on Github gist.
Training
- Training for YoloV5 6.2
Train custom models for YOLOv5 v6.2 with support for CPUs, CUDA enabled GPUs, and Apple Silicon.
v1.7.0 All Platforms except Raspberry Pi, Orange Pi, Radxa ROCK, Jetson Python, PyTorch, YOLO
Project by Matthew Dennis, based on Ultralytics YOLOv5.