Event-Based Crossing Dataset (EBCD)

Citation Author(s):
Riadul
Islam
University of Maryland, Baltimore County
Ryan
Robucci
University of Maryland, Baltimore County
Dhandeep
Challagundla
University of Maryland, Baltimore County
Joey
Mule
University of Maryland, Baltimore County
Rachit
Saini
University of Maryland, Baltimore County
Submitted by:
Dhandeep Challa...
Last updated:
Thu, 03/13/2025 - 13:21
DOI:
10.21227/ahdq-g045
Data Format:
License:
0
0 ratings - Please login to submit your rating.

Abstract 

Event-based vision revolutionizes traditional image sensing by capturing asynchronous intensity variations rather than static frames, enabling ultrafast temporal resolution, sparse data encoding, and enhanced motion perception. While this paradigm offers significant advantages, conventional event-based datasets impose a fixed thresholding constraint to determine pixel activations, severely limiting adaptability to real-world environmental fluctuations. Lower thresholds retain finer details but introduce pervasive noise, whereas higher thresholds suppress extraneous activations at the expense of crucial object information. To mitigate these constraints, we introduce the Event-Based Crossing Dataset (EBCD), a comprehensive dataset tailored for pedestrian and vehicle detection in dynamic outdoor environments, incorporating a multi-thresholding framework to refine event representations. By capturing event-based images at ten distinct threshold levels (4, 8, 12, 16, 20, 30, 40, 50, 60, and 75), this dataset facilitates an extensive assessment of object detection performance under varying conditions of sparsity and noise suppression. We benchmark state-of-the-art detection architectures—including YOLOv4, YOLOv7, EfficientDet-b0, MobileNet-v1, and Histogram of Oriented Gradients (HOG)—to experiment upon the nuanced impact of threshold selection on detection performance. By offering a systematic approach to threshold variation, we foresee that EBCD fosters a more adaptive evaluation of event-based object detection, aligning diverse neuromorphic vision with real-world scene dynamics. We present the dataset as publicly available to propel further advancements in low-latency, high-fidelity neuromorphic imaging.

Instructions: 

Event_Based_Crossing_Dataset.zip

 

Event_Based_Crossing_Dataset

 

threshold_4.zip

–test: 305 .txt and 305 .jpg files

–train: 2127 .txt and 2127 .jpg files

–valid: 607 .txt and 607 .jpg files

 

threshold_8.zip

–test: 305 .txt and 305 .jpg files

–train: 2127 .txt and 2127 .jpg files

–valid: 607 .txt and 607 .jpg files

 

threshold_12.zip

–test: 305 .txt and 305 .jpg files

–train: 2127 .txt and 2127 .jpg files

–valid: 607 .txt and 607 .jpg files

 

threshold_16.zip

–test: 305 .txt and 305 .jpg files

–train: 2127 .txt and 2127 .jpg files

–valid: 607 .txt and 607 .jpg files

 

threshold_20.zip

–test: 305 .txt and 305 .jpg files

–train: 2127 .txt and 2127 .jpg files

–valid: 607 .txt and 607 .jpg files

 

threshold_30.zip

–test: 305 .txt and 305 .jpg files

–train: 2127 .txt and 2127 .jpg files

–valid: 607 .txt and 607 .jpg files

 

threshold_40.zip

–test: 305 .txt and 305 .jpg files

–train: 2127 .txt and 2127 .jpg files

–valid: 607 .txt and 607 .jpg files

 

threshold_50.zip

–test: 305 .txt and 305 .jpg files

–train: 2127 .txt and 2127 .jpg files

–valid: 607 .txt and 607 .jpg files

 

 

threshold_60.zip

–test: 305 .txt and 305 .jpg files

–train: 2127 .txt and 2127 .jpg files

–valid: 607 .txt and 607 .jpg files

 

 

threshold_75.zip

–test: 305 .txt and 305 .jpg files

–train: 2127 .txt and 2127 .jpg files

–valid: 607 .txt and 607 .jpg files