Robust MLflow Deployment: From Localhost to Production
A practical guide to upgrading MLOps infrastructure by dockerizing MLflow, decoupling storage with PostgreSQL and MinIO, and performing a migration from local SQLite to a production-ready setup.
Computer Vision Lab
A Robotics, Computer Vision and Machine Learning lab by Nikolay Falaleev. The main focus of the blog is the application of Deep Learning for Computer Vision tasks, as well as other relevant topics: classical Computer Vision, Numerical Methods, and Hardware.
Latest posts
Benchmarking Nvidia RTX 5090
Benchmarking the top consumer-grade Nvidia Blackwell GPU against previous generations in Computer Vision and LLM applications.
Performance Analysis of Intel iGPUs in VLM and LLM applications
Analysis of cost-effective use of Intel iGPUs with ipex-llm to accelerate VLMs and LLMs with detailed benchmarks and practical setup instructions.
Multicamera systems: Calibration and beyond
Classical strategies and learning-based approaches for achieving accurate multicamera calibration in real-world.
Camera Calibration: Practical OpenCV Refinement Techniques
Practical advice on achieving better calibration accuracy in real-world scenarios while addressing common pitfalls and OpenCV model selection strategies.
Camera Calibration: What to perfect before touching the code
Physical and optical considerations of accurate real-world camera calibration - practical tips and techniques.
Camera calibration: Homography vs OpenCV
Comparison of the direct OpenCV camera calibration method with an approach based on a homography matrix.
Deep Learning in Sports and Autonomous Vehicles
Application of Deep Learning in Sports and similarities with the Self-Drivind field.
Top-1 solution of SoccerNet Camera Calibration Challenge 2023
An approach to generate camera calibration values from football broadcast videos.
Four generations of Nvidia GPUs compared
Benchmark results of GTX 1080 TI, RTX 2080Ti, 3090 and 4090 on DL tasks.
Multi-task learning loss balancing
Balancing temporal and spatial losses for simultenious training of multi-task neural networks for video processing and some data tips.
Benchmarking RTX 2080 Ti vs Pascal GPUs vs Tesla V100 with DL tasks
Comparation of Nvidia RTX 2080 Ti with GTX 1080 Ti and 1070.
Semantic Segmentation of Seismic Reflection Images
A gold medal solution of the TGS Salt Identification Challenge.
About Kaggle TGS Salt Identification Challenge
Kaggle Challenge to segment salt deposits beneath the Earth's surface on seismic images.
Discussion of the Lyft Perception Challenge
How to increase inference speed on a semantic segmentation task and further ideas.
Multiclass semantic segmentation with LinkNet34
A CNN approach used for multiclass semantic segmentation during the Lyft Perception Challenge.
About Lyft Perception Challenge
A partnered Lyft and Udacity semantic segmentation challenge with synthetic images.
An approach to predicted class balancing
An approach how to increase your position on a leaderpoard in a classification datascience competition by balancing predictions.
Camera model identification with deep learning
Ideas and approach to the Kaggle IEEE's Signal Processing Society - Camera Model Identification challenge.
How to run TensorFlow with CUDA 9 and cuDNN 7 in openSUSE on Ryzen
A set of instruction to run a modern version of the deep learning framework TensorFlow on AMD Ryzen.
Neural network for multiclass image segmentation
VGG16-based fully convolutional networks for semantic segmentation of images on Cityscapes
Image classification using SVM
Application of a linear SVM for image classification with HOG, binned color and color histogram features.
Adaptive Virtual Sensors for Lane detection
Computer vision approach for road marking detection with adaptive thresholds and positions of virtual sensors.
Bird's Eye View Transformation
Convert a normal image to a Bird's Eye view projection with OpenCV.
Camera calibration with OpenCV
A simple way to calibrate an optical system with a chessboard pattern by means of the OpenCV to reduce distortion.
Approximate equidistant curve for polynomials
An approximate approach to find an equidistant curve for a given polynomial.