"A neural network to rule them all, a neural network to find them, a neural network to bring them all and verify if is you !!" (Face recognition tool)

GPU-accelerated N-Body particle simulator with visualizer.

Because of the COVID-19 pandemic of 2020, more and more people are concerned with protecting themselves using masks, thus the need of software capable of monitoring whether the people are wearing masks or not. That is why I created a Python application using OpenCV (with CUDA support) based on the YOLOv4 algorithm, capable of monitoring the safe…

Media library face recognition tools

Stereo camera depth estimation with opencv and visualization in Open3D on jetson nano with CUDA support

Because of the COVID-19 pandemic of 2020, more and more people are concerned with protecting themselves using masks, thus the need of software capable of monitoring whether the people are wearing masks or not. That is why I created a Python application using OpenCV (with CUDA support) based on the YOLOv4 algorithm, more precisely the tiny option…

Docker implementation of the Marker pdf to markdown

Docker implementation of the Surya OCR

Real-Time ASR with CNN-BiLSTM: End-to-End Live Streaming Using PyTorch Lightning⚡

Docker implementation of the paper "Watermark Anything with Localized Messages"

An Ansible role to install the minimum needed to run a cuda application.

FracNeuPK: Fractional Neural Pharmacokinetics Simulator

Docker implementation of the Tabled OCR

Docker implementation of the EasyOCR

Docker implementation of the VideoSeal

A comprehensive pipeline for processing learning videos, generating subtitles, summarizing content, and creating structured documents.

Docker implementation of the Texify OCR

YoloV6 Onnxruntime Inference Pack

Intention is a configuration driven Youtube\Twitch downloader, sound file diarizer, transcriber and automatic host match and labeler. It also contains auto-topic classifiers and bias detection.

This repository presents a refined video summarization and sentiment analysis pipeline built entirely on video inputs using official Hugging Face models, including BLIP, SmolVLM (2.2B & 500M), XCLIP, and VideoMAE.