Readspeaker menu

NICE

New Image C Extension

Description

NICE is a C++-library for high level computer vision tasks.

It is built from different modules, where every module is optional and comes with its own Git repository.

Ressources

How to setup the library

Step 1: Obtain our main library
git clone https://github.com/cvjena/nice-core.git

Step 2: configure everything
cd nice-core/
source setenv.sh

Step 3: Obtain the disered sub-modules (e.g., gp-hik-core)
git clone https://github.com/cvjena/gp-hik-core.git

Step 4: Build everything
cd gp-hik-core/
make

Step 5: Verify that everything works properly
make check

Tutorials


Optional Sub-modules

gp-hik-core

Fast Gaussian process inference in large-scale applications. Project page on GitHub

gp-hik-exp (*)

Interfaces to integrate gp-hik-core into the vislearning extension. Comes along with many routines for our ECCV and ACCV experiments.

vislearning (*)

Sub-Library für High-Level Applications in the area of Image Understanding

  •  Regression
  • Classification
  •  Semantic Segmentation
  • High-Level Features
  • Incremental Learning
  • Novelty Detection
  • Active Learning


optimization (*)

Advanced optimization techniques.

limun (*)

3D computer vision, including matching, calibration, reconstruction and tracking.

(*) to be released on GitHub