Benchmark for evaluating semantic segmentation and object detection methods in maritime environment.

Dataset for defect-detection in industrial surfaces

Dataset for defect-detection in industrial surfaces

Traffic sign dataset cosisting of 200 categories in over 7000 images

This dataset contains marine videos, captured by unmanned surface vehicle (USV), for segmentation in three natural regions.

This dataset contains pixel-wise annotated maritime images, captured by unmanned surface vehicle (USV).

Dataset for video-based classification of combat positions in Jiu-jitsu

The dataset consists of 17 2D maps of indoor office and living environments generated from a subset of the Gibson Dataset.

Dataset consists of 612 reference images of an urban environment covering an area of 200 x 200 meters.

The dataset consists of 1572 geo-referenced images of an urban environment covering an area of 360 x 360 square meters.

Dataset consists of 1284 geo-referenced images of an urban environment covering an area of 800 x 200 square meters.

The dataset consists of 971 images of common fruit.

The dataset contains about 360 images of tree barks of 12 different trees.

Toolkit to support visual tracking performance evaluation

Schedule and run repeatable taks in Conda environments.

PyTorch implementation of SegDec-Net using weakly, mixed and fully supervised learning for surface defect detection. Implementation from ICPR2020 and COMIND2021 papers.

TensorFlow implementation of SegDec-Net for sufrace defect detection using deep neural networks. Implementation from JIM2019 paper.

TensorFlow implementation of displaced aggregation units for deep neural networks. Implementation from CVPR2018 and IJCV2020 papers.

Infinite data streams for deep learning.

Fork of the Detectron with added modifications for traffic sign detection. Implementation from TITS2020 paper.

Tracking Exchange protocol reference implementation

A simple and efficient interprocess communication (IPC) library for Linux written in C++11.

A Matlab toolbox for online multivariate kernel density estimation with unlearning routines and Gaussian mixture models compression algorithms.

A fast kernel density multivariate bandwidth estimation method.

A kernel density bandwidth estimation method tailored for 2D estimation problems.

A reference implementation of the unscented hellinger distance between two Gaussian mixture models.

A reference Matlab implementation of the Bayes spectral entropy measure of camera focus.