Hi there, I'm Mr. Kirosama 👋

🧠 Computer Vision Expert at Sony AI

🔬 Focused on attention mechanisms and image segmentation
🍣 AI and sushi lover!
🗾 Based in the state of Japan

💻 Skills and Interests

Computer Vision
Deep Learning
Attention Mechanisms
Image Segmentation
AI Research

🚀 Current Projects

Exploring advanced applications of Meta Transfer Learning for Few Shot Semantic Segmentation
Developing custom object detection models for real-time tracking
Enhancing Natural Language Processing for Serbian language
Extending Dual Attention VNet3D models for medical image segmentation
Creating advanced metrics for semantic segmentation evaluation

📊 GitHub Stats

🛠️ Languages and Tools

📫 How to reach me

GitHub: @kirosama2

Feel free to explore my repositories and don't hesitate to reach out for collaborations or discussions on AI and computer vision!

⭐️ From kirosama2

mr. kirosama's Projects

advanced-mtl-segmentation

An advanced application of Meta Transfer Learning for Few Shot Semantic Segmentation using U-Net framework.

advanced-semantic-segmentation-metrics

Advanced open-source tool for evaluating various metrics (miou, fiou, dice, accu, precision, recall) of semantic segmentation networks with multi-scale evaluation and crf post-processing capabilities.

extended-dual-attention-vnet3d

Extended DualAttention VNet3D Model for the KiTS19, an open-source project tackling challenges in Kidney Tumor Segmentation.

kirosama2-custom-detection-sagemaker

An open-source project demonstrating how to configure a custom object detection model and track the presence of certain objects within a defined session time. Perfect for building computer vision based services for a multitude of applications such as attendance tracking or security enforcement.

srbai-enhanced

An enhanced Python library for Serbian Natural Language Processing(NLP).

tensorrt-yolo-jetson-enhanced

This repository contains a revised version of the 'Jetson Nano: Deep Learning Inference Benchmarks Instructions', offering real-time object detection on Jetson Nano with a TensorRT optimized YOLO network.

Recommend Projects