解决Sim2Real中的数据不平衡:ImbalSim2Real方案及其应用在软机器人辅助康复的手指关节硬度自传感中
Zhongchao Zhou1, Yuxi Lu1, Pablo Enrique Tortós1
1Department of Medical System Engineering, Chiba University, Chiba, Japan.
Frontiers in bioengineering and biotechnology
|July 1, 2024
概括
不平衡的Sim2Real方案 (ImbalSim2Real) 通过增强区分器能力和确保定量映射来解决现实应用中的数据稀缺问题. 这种方法在sim2real任务中显著提高了模型性能,超过了现有的方法.
科学领域:
- 机器人技术 机器人技术 机器人技术
- 机器学习 机器学习
- 计算机视觉 计算机视觉
背景情况:
- 模拟与现实 (sim2real) 的差距给在模拟中训练的模型部署到现实世界的场景带来了挑战.
- 数据不平衡,现实数据稀缺,限制了传统sim2real方法的有效性,如CycleGAN.
- CycleGAN的局限性包括较低的区分能力和不确定的sim2real映射,在不平衡的数据集中尤其存在问题.
研究的目的:
- 提出一种新的不平衡的Sim2Real方案 (ImbalSim2Real),以克服现有方法在数据稀缺的sim2real场景中的局限性.
- 通过结合歧视者增强样本来增强歧视者的处理不平衡数据的能力.
- 通过整合对配对数据的目标回归损失来确保特定和定量映射.
主要方法:
- ImbalSim2Real方案采用双重训练方法,将数据细分为配对和不配对的集合.
- 未配对的数据被处理成分辨器增强样本,以提高分辨器的性能.
- 配对数据以目标回归损失进行训练,以完善生成器的映射精度.
主要成果:
- 数值实验验证实了 ImbalSim2Real 方案的优越性,而不是传统的 sim2real 方法.
- 在一个手指关节硬的自传感框架中,ImbalSim2Real与稀缺的现实数据的监督学习相比,减少了大约41%的验证损失.
- 与在不平衡数据集上训练的CycleGAN相比,拟议方案实现了验证损失的56%减少.
结论:
- 不平衡Sim2Real方案有效地解决了sim2real问题,特别是在数据不平衡严重的场景中.
- 拟议的框架证明了在生物信号估计任务中的实际适用性.
- ImbalSim2Real提供了一个强大的解决方案,可以弥合sim2real的差距,提高模型的概括性和性能.
相关概念视频
00:49Rigid Body Equilibrium Problems - I
A rigid body is said to be in static equilibrium when the net force and the net torque acting on the system is equal to zero. To solve for rigid body equilibrium problems, do the following steps.
01:21Rigid Body Equilibrium Problems - II
A rigid body is in static equilibrium when the net force and the net torque acting on the system are equal to zero.
Consider two children sitting on a seesaw, which has negligible mass. The first child has a mass (m1) of 26 kg and sits at point A, which is 1.6 meters (r1) from the pivot point B; the second child has a mass (m2) of 32 kg and sits at point C. How far from the pivot point B should the second child sit (r2) to balance the seesaw?
Consider two children sitting on a seesaw, which has negligible mass. The first child has a mass (m1) of 26 kg and sits at point A, which is 1.6 meters (r1) from the pivot point B; the second child has a mass (m2) of 32 kg and sits at point C. How far from the pivot point B should the second child sit (r2) to balance the seesaw?