Differentiable Computer Vision Library

Kornia — http://www.kornia.org & https://github.com/kornia/kornia — is a differentiable computer vision library for PyTorch.

It consists of a set of routines and differentiable modules to solve generic computer vision problems. At its core, the package uses PyTorch as its main backend both for efficiency and to take advantage of the reverse-mode auto-differentiation to define and compute the gradient of complex functions.
Overview

Inspired by OpenCV, this library is composed by a subset of packages containing operators that can be inserted within neural networks to train models to perform image transformations, epipolar geometry, depth estimation, and low-level image processing such as filtering and edge detection that operate directly on tensors.

The project is completely developed based on the community demand by an altruist and constantly growing developers team. In case you want to become a sponsor or a collaborator of the project, please feel to contact us. For sponsors, you can donate through a charitable tax-deductible stipend via the Open Source Vision Foundation (OSVF).

For further information, please contact us at [email protected]

We are in the process of planning our first 2019 Open Source Vision Foundation Conference. Stay tuned for more details.

More information @ http://learnopencv.com

The Driving Challenge

CARLA has been developed from the ground up to support development, training, and validation of autonomous urban driving systems. In addition to open-source code and protocols, CARLA provides open digital assets (urban layouts, buildings, vehicles) that were created for this purpose and can be used freely. The simulation platform supports flexible specification of sensor suites and environmental conditions.

For further information, please contact us at: [email protected]

More information @ http://opencv.org

A Modern Library for 3D Data Processing

Open3D — http://www.open3d.org & https://github.com/IntelVCL/Open3D — is a modern and easy-to-use library for 3D data processing. The availability of 3D data has increased significantly in recent years, but processing such data can be quite cumbersome. Open3D aims to make it easy. The project is now part of the non-profit Open Source Vision Foundation (OSVF, www.osvf.org) and is actively developed and maintained by a core team of developers in collaboration with the open source community. Our main objective is to continue to expand the functionality of Open3D so as to provide developers with efficient and easy to use tools for handling 3D. We aim to enable easy setup and maintenance of 3D-enabled pipelines in robotics, computer vision, machine learning, and autonomous driving applications.

Open3D has been built from scratch based on the design principles of usefulness and ease-of-use. Some of the main characteristics of Open3D are:

• Efficient C++11 back-end

• Modern Python bindings for fast prototyping

• Fast algebraic engine based on Eigen

• Compatibility with multiple standard point cloud formats

• Fast 3D data ingestion and processing

• Advanced and customizable visualization for rendering and interacting with 3D data

• State-of-the-art 3D processing algorithms for camera tracking, odometry, point cloud registration, pose-graph optimization, and scene reconstruction

Our team is actively working to bring a whole new set of features to enable the transparent use of 3D data in critical workflows such as

• Machine learning (training & inference)

• Geometric pipelines for SLAM, tracking, etc

• Data fusion in autonomous driving

• Manipulation in robotics

If you find these features useful or wish to suggest new features, please consider becoming a sponsor. Our sponsors play an active part in our road-map planning and prioritization. We ask sponsors to assist the development of Open3D through a charitable tax-deductible donation via the Open Source Vision Foundation (OSVF). The financial support can be granted on a one-time basis or periodically. We ask core sponsors to contribute $100,000 USD to support the development.

For further information, please contact us at [email protected]

The Silicon Valley Deep Learning Group™ provides education and career opportunities for technologists, entrepreneurs, investors, businessperson and companies in deep learning.

More information @ http://svdlg.com