phamquiluan/ResidualMaskingNetwork

ICPR 2020: Facial Expression Recognition using Residual Masking Network

What it solves

This project provides a state-of-the-art approach to Facial Expression Recognition (FER) and emotion detection, aiming to improve accuracy in identifying human emotions from images and video feeds.

How it works

The project implements a Residual Masking Network (RMN), a specialized neural network architecture designed for emotion detection. It supports training on standard datasets like FER2013 and ImageNet, and provides tools for inference on single frames or live webcam feeds. For higher accuracy, it supports an ensemble method that fuses the predictions of seven different models using a no-weighted sum average.

Who it’s for

Developers and researchers interested in computer vision and emotion AI, as well as those looking to integrate emotion detection into applications via a simple Python API or a client-side browser demo.

Highlights

  • High Accuracy: Outperforms several baseline models (like VGG19 and ResNet) on the FER2013 dataset.
  • Easy Integration: Installable via pip (pip install rmn) with a simple API for image and video detection.
  • Privacy-focused Demo: Includes a Hugging Face Space demo where inference runs entirely client-side using ONNX Runtime Web.
  • Comprehensive Benchmarking: Provides detailed accuracy results across FER2013 and ImageNet datasets.

Related

  • Project
    cosanlab/py-feat

    A Python toolbox for facial expression research that detects faces and extracts emotional expressions, facial muscle movements, and landmarks from images and videos.

  • Project
    dphnAI/sonar

    An inference engine for Hugging Face-compatible language and multimodal models that provides continuous batching, paged KV-cache management, and distributed serving.

  • Project
    lynote-ai/ai-image-detector

    AI Image Detector is an open‑source Python tool that estimates how likely an image was generated by AI. It ships with several ready‑made back‑ends (UnivFD, Sentry, Nonescape, hybrid ensembles) and lets you run a single command (`aidetect detect …`) on a file or folder, outputting a probability and label. Optional extras give a Gradio web UI or a FastAPI server. The repo includes benchmarking scripts, threshold‑calibration, and clear guidance on which backend to pick. It’s a lightweight, install‑once‑run‑anywhere utility for AI‑image detection, with documented limitations (no universal guarantee, no region‑level localization).

  • Project
    tomas-gajarsky/facetorch

    A Python library for facial detection and analysis that curates and optimizes open-source deep learning models for tasks like emotion recognition and deepfake detection.

  • Dispatch
    Hugging Face Accelerated Inference API Optimization

    Hugging Face achieved a 100x speedup in transformer inference for its Accelerated Inference API by combining high-level library optimizations, Rust-based tokenization, and hardware-specific compilation.