Deep learning-based object pose estimators are often unreliable and overconfident especially when the input image is outside the training domain, for instance, with sim2real transfer. Efficient and robust uncertainty quantification (UQ) in pose estimators is critically needed in many robotic tasks. In this work, we propose a simple, efficient, and plug-and-play UQ method for 6-DoF object pose estimation. We ensemble 2-3 pre-trained models with different neural network architectures and/or training data sources, and compute their average pairwise disagreement against one another to obtain the uncertainty quantification. We propose four disagreement metrics, including a learned metric, and show that the average distance (ADD) is the best learning-free metric and it is only slightly worse than the learned metric, which requires labeled target data. Our method has several advantages compared to the prior art: 1) our method does not require any modification of the training process or the model inputs; and 2) it needs only one forward pass for each model. We evaluate the proposed UQ method on three tasks where our uncertainty quantification yields much stronger correlations with pose estimation errors than the baselines. Moreover, in a real robot grasping task, our method increases the grasping success rate from 35% to 90%.
翻译:深深学习对象的估测值往往不可靠和过于自信, 特别是当输入图像在培训领域之外时, 例如, 模拟转移 。 许多机器人任务都非常需要高效力和强力的假估测值的不确定性量化( UQ ) 。 在这项工作中, 我们为 6- DoF 对象提出了简单、 高效、 插插插和游戏的 UQ 方法 进行估计 。 我们用不同的神经网络架构和/或培训数据源来混合2-3 预培训模型, 并计算出它们彼此之间的平均对对比, 以获得不确定性的量化。 我们建议了四个差异度指标, 包括一个学习的衡量标准, 并显示平均距离( UQ) 是最佳的无学习指标, 并且比学得的衡量标准略微差一些, 需要标注目标数据。 我们的方法与前一款相比有几个优点:(1) 我们的方法不需要对培训进程或模型输入作任何修改; 和(2) 我们只需要对每个模型做一个前导。 我们评估了三个任务的拟议UQ方法, 我们的不确定性量化方法在其中的三项任务中, 我们的不确定性量化能产生更强烈的关联性关系, 与35 和理解的正确度比理解的概率率率率 。