e-motion
收藏资源简介：
E-MOTION 是一个用于基于事件的场景流估计与独立运动物体研究的大规模数据集。该数据集旨在为计算机视觉，特别是事件相机和动态场景理解领域提供基准数据。数据内容主要包括一系列用于训练和测试的序列。训练序列包含多种运动模式，如平移（transl）、旋转（rot）、随机运动（rand）、方形轨迹（square_traj）、静态与动态的无人机（drone）、多物体场景（mult）以及多个静态与动态的日常物体（如漂白剂瓶、芥末瓶、薯片罐、饼干盒、水壶）。测试序列则涵盖了随机运动、方形轨迹、单/多无人机等场景。此外，数据集还提供了独立的 3D 模型文件，存放在 models 文件夹中。数据规模超过 1TB。该数据集适用于事件相机场景流估计、独立运动物体分割与跟踪、动态场景理解等任务。数据采用 ODbL v1.0 许可证发布，允许在注明出处且保持相同开放许可的前提下自由使用、修改和分享。
E-MOTION is a large-scale dataset for event-based scene flow estimation and independent moving object research. It aims to provide benchmark data for computer vision, particularly in the fields of event cameras and dynamic scene understanding. The data content mainly includes a series of sequences for training and testing. Training sequences cover various motion patterns, such as translation (transl), rotation (rot), random motion (rand), square trajectory (square_traj), static and dynamic drones (drone), multi-object scenes (mult), and multiple static and dynamic everyday objects (e.g., bleach bottle, mustard bottle, chips can, cookie box, kettle). Test sequences include scenarios like random motion, square trajectory, and single/multiple drones. Additionally, the dataset provides independent 3D model files stored in the models folder. The dataset size exceeds 1TB. It is suitable for tasks such as event camera scene flow estimation, independent moving object segmentation and tracking, and dynamic scene understanding. The data is released under the ODbL v1.0 license, allowing free use, modification, and sharing with attribution and under the same open license.
数据集名称
E-MOTION
许可协议
ODbL v1.0 许可协议，允许自由共享、修改和使用，但需注明出处，且修改后的数据库需保持开放并在相同许可下发布。
数据集规模
数据文件总量超过 1TB。
数据集描述
E-MOTION 是一个用于基于事件的独立运动物体场景流估计的数据集。
数据内容
-
测试序列：
- rand_slow_2
- rand_normal_1
- rand_fast_1
- square_traj_1
- drone_slow_1
- drone_normal_1
- drone_fast_2
- two_drones_1
- mult_2
-
训练序列：
- transl_slow_1, transl_slow_2, transl_normal_1, transl_normal_2, transl_fast_1
- rot_slow_1, rot_slow_2, rot_normal_1, rot_normal_2, rot_fast_1
- rand_slow_1, rand_normal_2, rand_fast_2, square_traj_2
- drone_static, drone_slow_2, drone_normal_2, drone_fast_1
- looping_static, looping_moving
- two_drones_2, square_traj_drone
- bleach_cleanser_static, bleach_cleanser
- mustard_bottle_static, mustard_bottle
- chips_can_static, chips_can
- cracker_box_static, cracker_box
- pitcher_static
- mult_static_1, mult_1, mult_static_2
-
3D 模型：位于
models文件夹中。
相关资源
- 数据集官方页面：https://heudiasyc.github.io/emotion/
- 数据自动下载脚本仓库：https://github.com/heudiasyc/emotion_utils.git
引用信息
如果您在学术背景中使用本数据集，请引用以下论文：
Rodriguez, I.G., Moreau, J., Bartolozzi, C., and Glover, A. E-MOTION: A Dataset for Event-Based Scene Flow Estimation with Independent Moving Objects, In European Conference on Computer Vision 2026.
bibtex @inproceedings{rodriguez2026emotion, title={E-MOTION: A Dataset for Event-Based Scene Flow Estimation with Independent Moving Objects}, author={Gutierrez Rodriguez, Ivan and Moreau, Julien and Bartolozzi, Chiara and Glover, Arren}, booktitle={European Conference on Computer Vision}, year={2026}, organization={Springer} }
联系方式
Iván Gutiérrez Rodríguez - igrodriguez@utc.fr
