• Home
  • Search
  • 이미지와 Voxel 을 이용한 3D 실내 점군 시맨틱 분할
  • https://doi.org/10.5302/j.icros.2021.21.0142Copy DOI Icon

이미지와 Voxel 을 이용한 3D 실내 점군 시맨틱 분할

Show More
  • Abstract
  • Literature Map
  • Similar Papers
Abstract

In this paper, we propose a parallel network architecture that exhibits improved performance by fusing two-dimensional (2D) and three-dimensional (3D) features. A voxel-based and a projection-based method were adopted to derive the results through one scan. Our approach consists of two parallel networks, extracts features along each dimension, and converges them in a fusion network. In the fusion network, the voxel blocks and 2D feature maps extracted from each structure are fused to the voxel grid and then trained through convolution. For effective training of 2D networks, we use data augmentation techniques using coordinate system rotation transformation. In addition, a multi-loss with weights applied to each dimension was employed to effectively enhance the performance of the system, and the results revealed that the system exhibited better performance than when a single loss was used. Our proposed method can achieve better performance by changing the performance of the 2D network and 3D network, which can be generalized using other structures.

Similar Papers
  • Research Article
  • Citations120

A hybrid deep segmentation network for fundus vessels via deep-learning framework

  • Mar 30, 2021
  • Neurocomputing
  • Lei Yang +4
  • Research Article
  • Citations7

Contribution and Marginal Effects of Landscape Patterns on Thermal Environment: A Study Based on the BRT Model

  • Aug 02, 2024
  • Buildings
  • Taojun Li +3
  • PDF
  • Research Article
  • Citations5

Improving cross-subject classification performance of motor imagery signals: a data augmentation-focused deep learning framework

  • Feb 02, 2024
  • Machine Learning: Science and Technology
  • Muhammed Enes Ozelbas +2
  • Conference Article

Decision fusion of 3D convolutional neural networks to triage patients with suspected prostate cancer using volumetric biparametric MRI

  • Mar 16, 2020
  • Medical Imaging 2020: Computer-Aided Diagnosis
  • Pritesh Mehta +5
  • Research Article
  • Citations5

Handwritten text generation and strikethrough characters augmentation

  • Jun 01, 2022
  • Computer Optics
  • A.V Shonenkov +5
  • Research Article
  • Citations11

Single image super resolution via wavelet transform fusion and SRFeat network

  • May 16, 2020
  • Journal of Ambient Intelligence and Humanized Computing
  • Chunyan Ma +5
  • Research Article

Feature visualization in large scale imaging mass spectrometry data

  • Jan 01, 2009
  • Data Archiving and Networked Services (DANS)
  • Alexander Broersen
  • Book Chapter
  • Citations52

Studying Robustness of Semantic Segmentation Under Domain Shift in Cardiac MRI

  • Jan 01, 2021
  • Peter M Full +3
  • Conference Article

Multilayer Subspace Learning With Self-Sparse Robustness for Two-Dimensional Feature Extraction

  • Jun 04, 2023
  • Han Zhang +3
  • Research Article
  • Citations8

Improving remote sensing scene classification using quality-based data augmentation

  • Mar 19, 2023
  • International Journal of Remote Sensing
  • Rowida Alharbi +4
  • Conference Article
  • Citations2

Block Scrambling Image Encryption Used in Combination with Data Augmentation for Privacy-Preserving DNNs

  • Sep 15, 2021
  • Tatsuya Chuman +1
  • Research Article

Deep learning and beamforming comparison of source ranging in a laboratory tank

  • Oct 01, 2023
  • The Journal of the Acoustical Society of America
  • Corey E Dobbs +1
  • Research Article
  • Citations8

Diabetic Retinopathy (DR) Image Synthesis Using DCGAN and Classification of DR Using Transfer Learning Approaches

  • Apr 29, 2023
  • International Journal of Image and Graphics
  • Yerrarapu Sravani Devi +1
  • Supplementary Content

Automated Embryo Selection Using Deep Learning: A Comparative Study of 3D-CNN, Hybrid, and Transfer Learning Models

  • Sep 04, 2025
  • Research Square
  • Aref Tavassoli Hojjati +4
  • Research Article
  • Citations5

A Low Slow Small Target Classification Network Model Based on K-Band Radar Dynamic Multifeature Data Fusion

  • Jan 01, 2025
  • IEEE Sensors Journal
  • Wang Yuan +5
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.