No Dense Tensors Needed: Fully Sparse Object Detection on Event-Camera Voxel Grids

• URL: http://arxiv.org/abs/2603.21638v1
• Date: Mon, 23 Mar 2026 07:12:14 GMT
• Title: No Dense Tensors Needed: Fully Sparse Object Detection on Event-Camera Voxel Grids
• Abstract summary: Event cameras produce asynchronous, high-dynamic-range streams well suited for detecting small, fast-moving drones.<n>Most event-based detectors convert the sparse event stream into dense tensors, discarding the representational efficiency of neuromorphic sensing.<n>We propose SparseVoxelDet, the first fully sparse object detector for event cameras, in which backbone feature extraction, feature pyramid fusion, and the detection head all operate exclusively on occupied voxel positions.
• Score: 0.3823356975862005
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Event cameras produce asynchronous, high-dynamic-range streams well suited for detecting small, fast-moving drones, yet most event-based detectors convert the sparse event stream into dense tensors, discarding the representational efficiency of neuromorphic sensing. We propose SparseVoxelDet, to our knowledge the first fully sparse object detector for event cameras, in which backbone feature extraction, feature pyramid fusion, and the detection head all operate exclusively on occupied voxel positions through 3D sparse convolutions; no dense feature tensor is instantiated at any stage of the pipeline. On the FRED benchmark (629,832 annotated frames), SparseVoxelDet achieves 83.38% mAP at 50 while processing only 14,900 active voxels per frame (0.23% of the T.H.W grid), compared to 409,600 pixels for the dense YOLOv11 baseline (87.68% mAP at 50). Relaxing the IoU threshold from 0.50 to 0.40 recovers mAP to 89.26%, indicating that the remaining accuracy gap is dominated by box regression precision rather than detection capability. The sparse representation yields 858 times GPU memory compression and 3,670 times storage reduction relative to the equivalent dense 3D voxel tensor, with data-structure size that scales with scene dynamics rather than sensor resolution. Error forensics across 119,459 test frames confirms that 71 percent of failures are localization near-misses rather than missed targets. These results demonstrate that native sparse processing is a viable paradigm for event-camera object detection, exploiting the structural sparsity of neuromorphic sensor data without requiring neuromorphic computing hardware, and providing a framework whose representation cost is governed by scene activity rather than pixel count, a property that becomes increasingly valuable as event cameras scale to higher resolutions.

Related papers

• YOLO12-MambaScan: An Efficient Object Detector with High-Frequency Enhancement and State-Space Modeling [3.6896352484416393]
  We propose an aerial-image detector built on the YOLO12 architecture.<n>The model combines a triple-path high-frequency enhancement convolution module, receptive-field coordinate-attention convolution (RFCAConv), and a Mamba-based global-context module.<n>On VisDrone, at an input resolution of 960*960, ours achieves 60.0% mAP@50 and 38.6%mAP@50:95, demonstrating a favorable accuracy--efficiency trade-off for small-object detection.
  arXiv  Detail & Related papers  (2026-09-12T01:58:35Z)
• An Event-Based Opto-Tactile Skin [6.8948204974942255]
  This paper presents a neuromorphic, event-driven tactile sensing system for soft, large-area skin.<n>It is based on the Dynamic Vision Sensors (DVS) integrated with a flexible silicone optical waveguide skin.<n>Such a design produces events as changes in brightness are detected, and estimates press positions on the 2D skin surface through triangulation.
  arXiv  Detail & Related papers  (2026-01-07T13:17:20Z)
• Interpolation-Based Event Visual Data Filtering Algorithms [0.0]
  We propose a method for event data that is capable of removing approximately 99% of noise while preserving the majority of the valid signal.<n>The proposed methods use about 30KB of memory for a sensor with a resolution of 1280 x 720 and is therefore well suited for implementation in embedded devices.
  arXiv  Detail & Related papers  (2025-07-02T10:13:20Z)
• SparseFormer: Detecting Objects in HRW Shots via Sparse Vision Transformer [62.11796778482088]
  We present a novel model-agnostic sparse vision transformer, dubbed SparseFormer, to bridge the gap of object detection between close-up and HRW shots.<n>The proposed SparseFormer selectively uses attentive tokens to scrutinize the sparsely distributed windows that may contain objects.<n> experiments on two HRW benchmarks, PANDA and DOTA-v1.0, demonstrate that the proposed SparseFormer significantly improves detection accuracy (up to 5.8%) and speed (up to 3x) over the state-of-the-art approaches.
  arXiv  Detail & Related papers  (2025-02-11T03:21:25Z)
• PLEIADES: Building Temporal Kernels with Orthogonal Polynomials [6.2921559513184695]
  We introduce a class of neural networks named PLEIA (Lynomial Expansion In Adaptive Distributed Event Systems), which contains temporal kernels generated from basis functions.<n>We focus on interfacing these networks with event-based data to perform online classification and detection with low latency.
  arXiv  Detail & Related papers  (2024-05-20T17:06:24Z)
• Dual Memory Aggregation Network for Event-Based Object Detection with Learnable Representation [79.02808071245634]
  Event-based cameras are bio-inspired sensors that capture brightness change of every pixel in an asynchronous manner. Event streams are divided into grids in the x-y-t coordinates for both positive and negative polarity, producing a set of pillars as 3D tensor representation. Long memory is encoded in the hidden state of adaptive convLSTMs while short memory is modeled by computing spatial-temporal correlation between event pillars.
  arXiv  Detail & Related papers  (2023-03-17T12:12:41Z)
• Fewer is More: Efficient Object Detection in Large Aerial Images [59.683235514193505]
  This paper presents an Objectness Activation Network (OAN) to help detectors focus on fewer patches but achieve more efficient inference and more accurate results. Using OAN, all five detectors acquire more than 30.0% speed-up on three large-scale aerial image datasets. We extend our OAN to driving-scene object detection and 4K video object detection, boosting the detection speed by 112.1% and 75.0%, respectively.
  arXiv  Detail & Related papers  (2022-12-26T12:49:47Z)
• Pushing the Limits of Asynchronous Graph-based Object Detection with Event Cameras [62.70541164894224]
  We introduce several architecture choices which allow us to scale the depth and complexity of such models while maintaining low computation. Our method runs 3.7 times faster than a dense graph neural network, taking only 8.4 ms per forward pass.
  arXiv  Detail & Related papers  (2022-11-22T15:14:20Z)
• SALISA: Saliency-based Input Sampling for Efficient Video Object Detection [58.22508131162269]
  We propose SALISA, a novel non-uniform SALiency-based Input SAmpling technique for video object detection. We show that SALISA significantly improves the detection of small objects.
  arXiv  Detail & Related papers  (2022-04-05T17:59:51Z)
• Operationalizing Convolutional Neural Network Architectures for Prohibited Object Detection in X-Ray Imagery [15.694880385913534]
  We explore the viability of two recent end-to-end object detection CNN architectures, Cascade R-CNN and FreeAnchor, for prohibited item detection. With fewer parameters and less training time, FreeAnchor achieves the highest detection inference speed of 13 fps (3.9 ms per image) The CNN models display substantial resilience to the lossy compression, resulting in only a 1.1% decrease in mAP at the JPEG compression level of 50.
  arXiv  Detail & Related papers  (2021-10-10T21:20:04Z)
• Small Object Detection Based on Modified FSSD and Model Compression [7.387639662781843]
  This paper proposes a small object detection algorithm based on FSSD. In order to reduce the computational cost and storage space, pruning is carried out to achieve model compression. The average accuracy (mAP) of the algorithm can reach 80.4% on PASCAL VOC and the speed is 59.5 FPS on GTX1080ti.
  arXiv  Detail & Related papers  (2021-08-24T03:20:32Z)
• ZoomNet: Part-Aware Adaptive Zooming Neural Network for 3D Object Detection [69.68263074432224]
  We present a novel framework named ZoomNet for stereo imagery-based 3D detection. The pipeline of ZoomNet begins with an ordinary 2D object detection model which is used to obtain pairs of left-right bounding boxes. To further exploit the abundant texture cues in RGB images for more accurate disparity estimation, we introduce a conceptually straight-forward module -- adaptive zooming.
  arXiv  Detail & Related papers  (2020-03-01T17:18:08Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.