GitHub

Flash Attention with Sink — GPT-OSS 20B Attention Implementation

flash-attention-with-sink implements an attention variant used in GPT-OSS 20B that integrates a "sink" step into FlashAttention. This repo focuses on the forward path and provides an experimental ...
IEEE

Implementation and Comparison of CNN Models on CIFAR-10 Dataset

Abstract: This study investigates the vulnerability of Convolutional Neural Network (CNN) models to adversarial attacks, focusing on the Fast Gradient Sign Method (FGSM). We implemented and compared ...
IEEE

Design and Implementation of a Flexible CNN Accelerator for Fast Real-Time Object Detection on FPGA

Abstract: In edge computing systems with limited resources, such as mobile devices and the Internet of Things, the use of Convolutional Neural Network (CNN) accelerators on FPGA has increasingly ...
GitHub

Neural Network from Scratch

Eric Gutiérrez, 6th February 2026. A Python implementation of a 1-hidden layer neural network built entirely from first principles. This project avoids deep learning libraries (like TensorFlow or ...
IEEE

Design and Implementation of Face Recognition System Based on CNN

Abstract: In the era of rapid development of information technology, facial recognition technology has seen widespread application in areas such as security surveillance, identity verification, and ...
IEEE

Implementation of an Ultra-Low Power CNN Accelerator on PYNQ Z2 FPGA-Based Platform for Real-Time Edge AI Applications

Abstract: In this paper, the design and implementation of an ultra-low power Convolutional Neural Network (CNN) accelerator on the Xilinx PYNQ-Z2 FPGA will be described. The system is an Edge AI ...
GitHub

Innoactive Portal Python Client

Command-line client for Innoactive Portal's APIs, written in Python. To authenticate against Portal Backend, you need to provide credentials as environment variables. You can use either a Bearer token ...
IEEE

FPGA Implementation of CNN-Based Pulmonary Disease Classifier for Low-Cost Diagnostic Platforms

Abstract: The paper introduces the design and FPGA implementation of a convolutional neural network (CNN)-based, low-cost and low-energy diagnostic platform that classifies pulmonary diseases. The ...