Point Virtual Transformer

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Sood, Veerain, Bnalin, Pandey, Gaurav
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917254261637120
author Sood, Veerain
Bnalin
Pandey, Gaurav
author_facet Sood, Veerain
Bnalin
Pandey, Gaurav
contents LiDAR-based 3D object detectors often struggle to detect far-field objects due to the sparsity of point clouds at long ranges, which limits the availability of reliable geometric cues. To address this, prior approaches augment LiDAR data with depth-completed virtual points derived from RGB images; however, directly incorporating all virtual points leads to increased computational cost and introduces challenges in effectively fusing real and virtual information. We present Point Virtual Transformer (PointViT), a transformer-based 3D object detection framework that jointly reasons over raw LiDAR points and selectively sampled virtual points. The framework examines multiple fusion strategies, ranging from early point-level fusion to BEV-based gated fusion, and analyses their trade-offs in terms of accuracy and efficiency. The fused point cloud is voxelized and encoded using sparse convolutions to form a BEV representation, from which a compact set of high-confidence object queries is initialised and refined through a transformer-based context aggregation module. Experiments on the KITTI benchmark report 91.16% 3D AP, 95.94% BEV AP, and 99.36% AP on the KITTI 2D detection benchmark for the Car class.
format Preprint
id arxiv_https___arxiv_org_abs_2602_06406
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Point Virtual Transformer
Sood, Veerain
Bnalin
Pandey, Gaurav
Computer Vision and Pattern Recognition
LiDAR-based 3D object detectors often struggle to detect far-field objects due to the sparsity of point clouds at long ranges, which limits the availability of reliable geometric cues. To address this, prior approaches augment LiDAR data with depth-completed virtual points derived from RGB images; however, directly incorporating all virtual points leads to increased computational cost and introduces challenges in effectively fusing real and virtual information. We present Point Virtual Transformer (PointViT), a transformer-based 3D object detection framework that jointly reasons over raw LiDAR points and selectively sampled virtual points. The framework examines multiple fusion strategies, ranging from early point-level fusion to BEV-based gated fusion, and analyses their trade-offs in terms of accuracy and efficiency. The fused point cloud is voxelized and encoded using sparse convolutions to form a BEV representation, from which a compact set of high-confidence object queries is initialised and refined through a transformer-based context aggregation module. Experiments on the KITTI benchmark report 91.16% 3D AP, 95.94% BEV AP, and 99.36% AP on the KITTI 2D detection benchmark for the Car class.
title Point Virtual Transformer
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2602.06406