Software and Models

We provide various models and software packages (including source codes) for downloads. A few stats. on our work as of 2018:

  • Refereed, international papers: About 80.
  • Total citations: About 4200 (Google scholar).
  • SW downloads: About 1k to 2k per year

We also give tutorials or talks on our SWs. Talk slides and other materials of those talks are too.


Open source projects

We try to release source codes of all of our research activities. As a result, there are too many of them. We list codes released before 2016.
Some of recent ones are at github yoonlab, github sglab or individual paper webpage.

Collision-Backpropagation based Obstacle Avoidance Method for a Legged Robot Expressed as a Simplified Dynamics Model

Contour-Aware Equipotential Learning for Semantic Segmentation

Diverse Generative Perturbations on Attention Space for Transferable Adversarial Attacks

Semi-Supervised Learning of Optical Flow by Flow Supervisor

Part-based Pseudo Label Refinement for Unsupervised Person Re-identification

Deep Video Inpainting Guided by Audio-Visual Self-Supervision

Analysis and Acceleration of TORM: Optimization-based Planning for Path-wise Inverse Kinematics

Implicit LiDAR Network: LiDAR Super-Resolution via Interpolation Weight Prediction

Confidence-Based Robot Navigation Under Sensor Occlusion with Deep Reinforcement Learning

RCIK: Real-Time Collision-Free Inverse Kinematics Using a Collision-Cost Prediction Network

In-N-Out: Towards Good Initialization for Inpainting and Outpainting

MeshChain: Secure 3D model and intellectual property management powered by blockchain technology

Weakly-Supervised Contrastive Learning in Path Manifold for Monte Carlo Image Reconstruction

GraphShop: Graph Approach for Shop-type Recommendation

Adaptive Kernel Inference for Dense and Sharp Occupancy Grids

Unsupervised Learning of Optical Flow with Deep Feature Similarity

An Objectness Score for Accurate and Fast Detection during Navigation

Harmonious Sampling for Mobile Manipulation Planning

Super Rays and Culling Region for Real-Time Updates on Grid-based Occupancy Maps

Scale-Varying Triplet Ranking with Classification Loss for Facial Age Estimation

Kinodynamic Comfort Trajectory Planning for Car-like Robots

Automated task planning using object arrangement optimization

Simultaneous Planning of Sampling and Optimization with Configuration Free Space Approximation

Reflection-Aware Sound Source Localization

Regional Attention Based Deep Feature for Image Retrieval

Image Completion with Intrinsic Reflectance Guidance

Ray Distribution to Parallel Batching-based Updates

  • ParallelBatch (2017): High performance method for updating tree-based occupancy maps using multi-threads

Rank-based Voting with Inclusion Relationship for Accurate Image Search

PCA based Computation of Illumination-Invariant Space

Physically Inspired, Interactive Lightening Generation

Anytime RRBT for Handling Uncertainty and Dynamic Objects

Super Ray based Updates for Occupancy Maps

IsFAM: Discriminative Subgraphs for Discovering Family Photos

  • IsFAM (2016): an open source project for discovering family photos using discriminative subgraphs.

IR-SKELETON: An educational solution framework for OpenGL and Instant Radiosity

Naive Bayes Nearest Neighbor

  • NBNN (2015): a library for image classification using NBNN.

Path/Motion Planning plugin for V-REP simulatorNaive Bayes Nearest Neighbor

Adaptive rendering for Monte Carlo ray tracing

Denoising method for Monte Carlo ray tracing

High-Quality Rendering for Large-Scale Models

  • T-Rex (2014): source codes of our T-Rex rendering system.

Binary Code Embedding Methods

Collision Detection

  • OpenCCD (2009): a open source project (github project) for performing continuous collision detection for deforming models.
    Its old web page is here

Fast rendering framework SDK

  • OpenIRT (2015):  a fast rendering framework SDK which supports various algorithms like
    • Whitted ray tracing, Path tracing, Photon mapping, T-ReX (, Rasterization (currently, OpenGL based flat shading)

Random-Accessible Compressed Data

  • OpenRACM (2007): a library for computing random-accessible compressed triangle meshes.

Cache-Coherent Layout Computation

  • OpenCCL 1.6 (2015): a open source project (github project) for computing cache-coherent layouts of meshes and graphs.
    Its old version (’05 version) is here