Dense 3D scene reconstruction from an ordered sequence or unordered image collections is a critical step when bringing research in computer vision into practical scenarios. Following the paradigm introduced by DUSt3R, which unifies an image pair densely into a shared coordinate system, subsequent methods maintain an implicit memory to achieve dense 3D reconstruction from more images. However, such implicit memory is limited in capacity and may suffer from information loss of earlier frames. We propose Point3R, an online framework targeting dense streaming 3D reconstruction. To be specific, we maintain an explicit spatial pointer memory directly associated with the 3D structure of the current scene. Each pointer in this memory is assigned a specific 3D position and aggregates scene information nearby in the global coordinate system into a changing spatial feature. Information extracted from the latest frame interacts explicitly with this pointer memory, enabling dense integration of the current observation into the global coordinate system. We design a 3D hierarchical position embedding to promote this interaction and design a simple yet effective fusion mechanism to ensure that our pointer memory is uniform and efficient. Our method achieves competitive or state-of-the-art performance on various tasks with low training costs.

Multi-view multi-person 3D pose estimation aims to localize 3D skeleton joints for each person instance in a scene from multi-view camera inputs. It is a fundamental task that benefits many real-world applications (such assurveillance, sportscast, gaming and mixed reality) and ismainly tackled byreconstruction-based [6,14,4]andvolumetric [40]approaches inpreviousliterature, as showninFig.1(a)and(b).

The proposed MLP decoder aggregates information from different layers, andthus combining both local attention and global attention to render powerful representations.

Compared toprevious works, our approach can learn landmarks that are more flexible in terms of capturing large changes in viewpoint.

The shared part contains information for both domains. The exclusive parts, on the other hand, contain only factors of variation that are particular to each domain.

However,most existing studies concentrate onthetext-to-image synthesis [11-14], which generates images from text descriptions without the original image.

Recent progress hasdemonstrated thatsuchmeta-learning methods may exceed scalable human-invented architectures on image classification tasks.

To tackle such an issue as well as promote the quality of object attention, we introduce asimple yet effectiveSelfErasing Network (SeeNet) to prohibit attentions from spreading to unexpected background regions.

Recent high-performing object detectors, such asFaster R-CNN [34],formulate thedetection problem as atwo-stage pipeline.

Such biases are known asrepresentation bias [38].