Cluttered bin-picking environments are challenging for pose estimation models. Despite the impressive progress enabled by deep learning, single-view RGB pose estimation models perform poorly in cluttered dynamic environments. Imbuing the rich temporal information contained in the video of scenes has the potential to enhance models ability to deal with the adverse effects of occlusion and the dynamic nature of the environments. Moreover, joint object detection and pose estimation models are better suited to leverage the co-dependent nature of the tasks for improving the accuracy of both tasks. To this end, we propose attention-based temporal fusion for multi-object 6D pose estimation that accumulates information across multiple frames of a video sequence. Our MOTPose method takes a sequence of images as input and performs joint object detection and pose estimation for all objects in one forward pass. It learns to aggregate both object embeddings and object parameters over multiple time steps using cross-attention-based fusion modules. We evaluate our method on the physically-realistic cluttered bin-picking dataset SynPick and the YCB-Video dataset and demonstrate improved pose estimation accuracy as well as better object detection accuracy

翻译：杂乱料箱拣选环境对姿态估计模型构成严峻挑战。尽管深度学习取得了显著进展，但单视角RGB姿态估计模型在杂乱动态环境中表现欠佳。利用视频场景中包含的丰富时序信息，有望增强模型应对遮挡及环境动态性不利影响的能力。此外，联合目标检测与姿态估计模型更能利用两类任务的相互依赖性，从而提升各自精度。为此，我们提出基于注意力机制的时序融合方法，用于多目标6D姿态估计，该方法可在视频序列的多帧间累积信息。我们的MOTPose方法以图像序列为输入，在一次前向传播中完成对所有目标的联合检测与姿态估计。通过基于交叉注意力的融合模块，该方法学习在多个时间步上聚合目标嵌入向量及目标参数。我们在物理逼真的杂乱料箱拣选数据集SynPick和YCB-Video数据集上进行了评估，结果表明该方法在提升姿态估计精度的同时，也实现了更优的目标检测准确率。