Securade.ai HUB - A generative AI based edge platform for computer vision that connects to existing CCTV cameras and makes them smart.

A lightweight, edge-first web framework built on Cloudflare Workers with authentication, D1 database, and a clean dark UI. Deploy a blog or small site globally in under 5 minutes. Version 1.0

Ingredient to Sugar Level Estimation (from training in Python to edge deployment in JS/TS)

Proyek ini menggunakan kerangka deteksi objek berbasis YOLO (You Only Look Once) untuk memantau ternak ayam secara real-time lewat kamera CCTV IP. Sistem kemudian mengintegrasikan hasil deteksi dengan komponen IoT (seperti kamera, pengiriman data via MQTT/HTTP, dan perangkat edge)

This project is an AI-powered mobile application capable of recognizing age, gender, and facial expressions from images.

Light-weight 6D pose estimation for Edge devices

POSIX-compliant configuration parser for systematic build coordination with deterministic pass-mode resolution. Phase 1 implementation establishing foundational architecture for modular component discovery, threading infrastructure, and systematic validation within the NexusLink ecosystem. Waterfall methodology with comprehensive quality assurance.

Real-time SAM2 segmentation on edge devices - 40x faster C++ inference with ONNX Runtime for iOS/Android deployment

Example app using React Create App & Digital Optimization Group's ADN & CMS

Optimized CNN achieving ~90% accuracy with 38.6% parameter reduction for production-ready digit recognition

Toolset for creating and publishing OS images with automated TPM attestation process for Azure IoT Edge.

Code of the paper "Emotion Recognition on Edge Devices: Training and Deployment " by Pandelea et al.

Task-adaptive pruning framework for deploying Vision Transformers on heterogeneous edge devices without accessing private data (arXiv 2601.02437)

YOLOv3-YOLO12 unified pipeline for edge deployment - Detection, segmentation, pose estimation with PyTorch to ONNX/TFLite/CoreML export

UVA DS 6050 final project. This aims to build smaller models that are easier to use on edge devices

YOLOE-Unified is a novel framework that integrates YOLOE with distilled CLIP, runtime SAM refinement, and TensorRT optimization for efficient open-vocabulary object detection and instance segmentation on edge devices (Jetson Orin, etc.).

edge AI model deployment toolkit for Google Coral TPU — Part 3 of 3 in the Philippine license plate recognition pipeline

MobileNetV3 object detection with TFLite quantization — fp32/fp16/int8 edge deployment benchmarks

Facility Health Assessment for Edge Deployment — MLP autoencoder + severity classifier, ~15K params, real-time monitoring, SLSA Level 3

Developing efficient deep learning models for real-world use. Covers knowledge distillation, quantization, pruning, and more. Focused on reducing size and latency while preserving accuracy. Includes training pipelines, visualizations, and performance reports.