ASIST: Automatic semantically invariant scene transformation

ASIST: Automatic semantically invariant scene transformation

We present ASIST, a technique for transforming point clouds by replacing objects with their semantically equivalent counterparts. Transformations of this kind have applications in virtual reality, repair of fused scans, and robotics. ASIST is based on a unified formulation of semantic labeling and object replacement; both result from minimizing a single objective. We present numerical tools for...

Automatic Camera Calibration and Scene Reconstruction with Scale-Invariant Features

The goal of our research is to robustly reconstruct general 3D scenes from 2D images, with application to automatic model generation in computer graphics and virtual reality. In this paper we aim at producing relatively dense and well-distributed 3D points which can subsequently be used to reconstruct the scene structure. We present novel camera calibration and scene reconstruction using scale-...

Semantically Driven Automatic Hyperlinking

Semantically Enhanced Automatic Keyphrase Indexing

The goal of this PhD thesis is to elaborate methods for automatic keyphrase indexing with a controlled vocabulary. Keyphrases are single words or multi-word lexemes that concisely and accurately describe the subject or an aspect of the subject discussed in a document. They are widely used in large document collections such as digital libraries and document repositories. They help organize mater...

Automatic Scene Activity Modeling

Motivation: A scene activity modeling system should be capable of modeling a new environment with minimal prior information or human intervention. Priors which depend on particular camera placements or particular types of trackable objects pigeon-hole an application to a certain type of task and make it vulnerable to unusual objects or activities. This work attempts to model all the significant...