Neural network machine learning approaches are widely used for object classification or detection problems with significant success. A similar problem with specific constraints and challenges is object state estimation, dealing with objects that consist of several removable or adjustable parts. A system that can detect the current state of such objects from camera images can be of great importance for Augmented Reality (AR) or robotic assembly and maintenance applications. In this work, we present a CNN that is able to detect and regress the pose of an object in multiple states. We then show how the output of this network can be used in an automatically generated AR scenario that provides step-by-step guidance to the user in assembling an object consisting of multiple components.

Monocular 3D object detection has recently made a significant leap forward thanks to the use of pre-trained depth estimators for pseudo-LiDAR recovery. Yet, such two-stage methods typically suffer from overfitting and are incapable of explicitly encapsulating the geometric relation between depth and object bounding box. To overcome this limitation, we instead propose to jointly estimate dense scene depth with depth-bounding box residuals and object bounding boxes, allowing a two-stream detection of 3D objects that harnesses both geometry and context information. Thereby, the geometry stream combines visible depth and depth-bounding box residuals to recover the object bounding box via explicit occlusion-aware optimization. In addition, a bounding box based geometry projection scheme is employed in an effort to enhance distance perception. The second stream, named as the Context Stream, directly regresses 3D object location and size. This novel two-stream representation enables us to enforce cross-stream consistency terms, which aligns the outputs of both streams, and further improves the overall performance. Extensive experiments on the public benchmark demonstrate that OPA-3D outperforms state-of-the-art methods on the main Car category, whilst keeping a real-time inference speed.

Estimation and tracking of 6DoF poses of objects in images is a challenging problem of great importance for robotic interaction and augmented reality. Recent approaches applying deep neural networks for pose estimation have shown encouraging results. However, most of them rely on training with real images of objects with severe limitations concerning ground truth pose acquisition, full coverage of possible poses, and training dataset scaling and generalization capability. This paper presents a novel approach using a Convolutional Neural Network (CNN) trained exclusively on single-channel Synthetic images of objects to regress 6DoF object Poses directly (SynPo-Net). The proposed SynPo-Net is a network architecture specifically designed for pose regression and a proposed domain adaptation scheme transforming real and synthetic images into an intermediate domain that is better fit for establishing correspondences. The extensive evaluation shows that our approach significantly outperforms the state-of-the-art using synthetic training in terms of both accuracy and speed. Our system can be used to estimate the 6DoF pose from a single frame, or be integrated into a tracking system to provide the initial pose.

Assembly state detection, i.e., object state detection, has a critical meaning in computer vision tasks, especially in AR assisted assembly. Unlike other object detection problems, the visual difference between different object states can be subtle. For the better learning of such subtle appearance difference, we proposed a two-level group attention module (TGA), which consists of inter-group attention and intro-group attention. The relationship between feature groups as well as the representation within each feature group is simultaneously enhanced. We embedded the proposed TGA module in a popular object detector and evaluated it on two new datasets related to object state estimation. The result shows that our proposed attention module outperforms the baseline attention module.

Estimating the 6D pose of an object from a single RGB image is a critical task that becomes additionally challenging when dealing with symmetric objects. Recent approaches typically establish one-to-one correspondences between image pixels and 3D object surface vertices. However, the utilization of one-to-one correspondences introduces ambiguity for symmetric objects. To address this, we propose SymCode, a symmetry-aware surface encoding that encodes the object surface vertices based on one-to-many correspondences, eliminating the problem of one-to-one correspondence ambiguity. We also introduce SymNet, a fast end-to-end network that directly regresses the 6D pose parameters without solving a PnP problem. We demonstrate faster runtime and comparable accuracy achieved by our method on the T-LESS and IC-BIN benchmarks of mostly symmetric objects. The code is available at https://github.com/lyltc1/SymNet.

We propose a finite element method (FEM) based approach for surface stitching which can be integrated into existing SLAM and NRSfM pipelines for AR applications. Given individual reconstructions and camera poses at different time stamps, our stitching method incrementally completes the surface with a smooth transition between the hidden and the observed parts, so that all the observed parts can be stitched into a single surface. Thanks to the physical modelling, deformations from the observed parts are propagated to the hidden parts enabling an overall high-fidelity and realistic estimate. To keep the computational time in bounds, deformations near the observed parts are computed with FEM, and the remaining region is approximated by Laplacian deformation. We assume that no force is applied to the hidden parts. To evaluate the algorithm, we generate a synthetic dataset with ground truth. In our dataset, the camera observes only a part of the target surface in each frame and moves until the whole target surface is covered. The dataset which will be made publicly available includes the ground truth camera poses and geometries of the whole surface at each time frame. An experimental evaluation of the stitching method with accuracy metrics rounds out the draft.