YCB benchmark dataset
收藏资源简介:
该数据集来源于YCB基准数据集,包含10个不同物体的主动探索数据,每个物体100个样本。数据集通过在QB软手上集成柔性本体感受传感器,使用肌肉梭数学模型将传感器的测量值实时编码为脉冲序列。这些脉冲序列随后被输入到一个混合脉冲神经网络进行分类。该数据集有助于研究本体感受信号在人工系统中的使用,尤其是在物体分类任务中。
This dataset is derived from the YCB benchmark dataset, containing active exploration data for 10 distinct objects, with 100 samples collected for each object. The dataset is built by integrating flexible proprioceptive sensors onto the QB soft hand, where a mathematical model of muscle spindles is used to encode the sensor measurements into real-time spike sequences. These spike sequences are then fed into a hybrid spiking neural network to perform classification tasks. This dataset facilitates research on the application of proprioceptive signals in artificial systems, especially in object classification tasks.

- 1Object Classification Utilizing Neuromorphic Proprioceptive Signals in Active Exploration: Validated on a Soft Anthropomorphic Hand慕尼黑工业大学认知系统研究所, 约翰霍普金斯大学生物医学工程系 · 2025年



