tensorflow/tfjs-models
Tfjs Models
This project is a collection of pre-trained machine learning models and conversion pipelines designed for running inference directly in the browser using TensorFlow.js. It provides a library of ready-to-use models for computer vision, audio classification, and natural language processing tasks.
The suite includes specialized tools for transforming Python-based Keras models into JSON formats compatible with web environments. It enables the deployment of these models by fetching architectures and weight shards via HTTP for client-side execution.
The project covers a broad range of capabilities, including image and object classification, human pose and landmark detection, and image depth estimation. It also supports text toxicity analysis, the generation of numerical embeddings for semantic similarity, and the categorization of short audio recordings and speech commands.
Additional utilities provide support for transfer-learning based classification, data preparation for multidimensional tensors, and the visualization of model behavior and training metrics.
Features
- Browser-Based Model Libraries - Provides a comprehensive library of ready-to-use machine learning models for direct browser-side inference.
- Browser-based Inference Engines - Executes pre-trained machine learning models directly within web browsers using TensorFlow.js.
- Computer Vision Models - Offers a suite of pre-trained models for image classification, object detection, human pose tracking, and depth estimation.
- Image Classification - Identifies visual labels and locates multiple objects within a frame through semantic segmentation.
- Web-Based Computer Vision - Identifies objects, detects human poses, and estimates image depth directly in a web browser.
- Model Conversion Pipelines - Transforms Python-based Keras models into JSON formats compatible with browser-based execution.
- Natural Language Processing - Analyzes text for toxicity or creates numerical embeddings for similarity tasks without sending data to a server.
- Natural Language Processing Analysis - Evaluates text for toxicity and converts natural language into numerical embeddings for similarity analysis.
- Client-Side Model Loading - Enables the deployment of models by fetching architectures and weight shards via HTTP for client-side execution.
- Audio Event Classification - Analyzes short audio recordings and speech commands within a browser to identify spoken words or sounds.
- Audio Classification Models - Provides pre-trained neural networks that identify spoken words and categorize short audio recordings in real time.
- Speech Command Classification - Classifies short audio recordings into specific categories using speech command datasets to identify spoken words.
- Model Conversion - Converts machine learning models trained in Python into formats compatible with browser execution.
- Keras - Transforms Python-based Keras models into JSON formats compatible with web browsers.
- TensorFlow - Converts and loads pretrained Python models into a browser for real time inference and evaluation.
- Tensor Processing Pipelines - Transforms raw pixels or audio samples into multidimensional arrays for parallel GPU execution within the model pipeline.
- Human Pose Detections - Tracks body positions, hand poses, and facial geometry to determine movement and surface structure in real time.
- Model Serialization - Stores model topology and weights in JSON format to enable loading over HTTP into a browser environment.
- Depth Estimation - Calculates the distance from the camera to the subject for every pixel to create 3D visual effects.
- ML Model Visualizations - Renders training metrics and model behavior in a web interface to analyze how data is processed.
- Weight Sharding - Splits large model weights into smaller binary files to optimize loading speed and memory usage during initialization.
- Text Embeddings - Provides local generation of numerical text embeddings to determine semantic similarity without server requests.
- Transfer Learning - Adapts pre-trained models to new categories by adding custom layers and training on smaller datasets.
- Deep Learning Frameworks - Collection of pre-trained models for TensorFlow.js.
Star history
How this analysis was created: This summary and feature list are AI-generated from collected project material and can contain mistakes. Stars, license and language are imported from GitHub. Inclusion does not mean that we have tested or audited this project. Check the source documentation for any feature you depend on. Learn more on our About page.
