Abstract:Although tremendous strides have been made in object recognition over large-scale 3D point clouds, one of the remaining open challenges is detecting tiny objects. The signal and appearance information of tiny objects are generally sparse and insufficient due to their tiny size, which makes the tiny object difficult to be detected with existing studies both in tiny 2D object recognition and general 3D object recognition since 3D topology relations are lost due to feature projection in tiny 2D object recognition and massive background may dominate the feature learner in general 3D object recognition. To the best of our knowl-edge, we explore tiny object recognition for the first time over large-scale 3D point clouds by designing deep neural network. To cope with this problem, a novel two-stage approach is proposed in this paper, which can effectively learn an informative and discriminative feature representation and efficiently pro-cess large-scale point clouds. Specifically, a class-aware filtering strategy is designed to dispense with the redundant background information. After that, we introduce a novel aggregating and sampling scheme in a supervised manner to progressively increase the receptive field for each 3D point, which can adaptively capture useful signals from tiny objects and learn complex geometric structures. Last, we propose a biased loss function that is inversely correlated to the number of tiny object points, thereby effectively handling the sparsity of tiny objects. Extensive experimental results over two real-world data sets vali-date the effectiveness and efficiency of the proposed approach.(c) 2023 Elsevier B.V. All rights reserved.

Small Object Recognition Using a Spatio-Temporal Neural Network

$\mathcal{R}^2$ -CNN: Fast Tiny Object Detection in Large-Scale Remote Sensing Images

A fast multi-scale convolutional neural network for object recognition

Sparse Temporal Encoding of Visual Features for Robust Object Recognition by Spiking Neurons

Spatial-Transformer and Cross-Scale Fusion Network (STCS-Net) for Small Object Detection in Remote Sensing Images

An Event-Driven Object Recognition Model Using Activated Connected Domain Detection

Small object detection based on attention mechanism and enhanced network

Class-aware tiny object recognition over large-scale 3D point clouds

Small Object Detection using Multi-scale Feature Fusion and Attention

Multi-scale receptive field neural networks for object tracking

ALFPN: Adaptive Learning Feature Pyramid Network for Small Object Detection

SSN: Scale Selection Network for Multi-Scale Object Detection in Remote Sensing Images

Small Object Detection Method Based on Adaptive Spatial Parallel Convolution and Fast Multi-Scale Fusion

Feature Rescaling and Fusion for Tiny Object Detection

Attention-based scale sequence network for small object detection

SizeNet: Object Recognition via Object Real Size-based Convolutional Networks

Multi-scale detector optimized for small target

An Evaluation of Deep Learning Methods for Small Object Detection

iSmallNet: Densely Nested Network with Label Decoupling for Infrared Small Target Detection

ScaleNAS: One-Shot Learning of Scale-Aware Representations for Visual Recognition

A Deep Lightweight Convolutional Neural Network Method for Real-Time Small Object Detection in Optical Remote Sensing Images