[PDF] Multi-view object class detection with a 3D geometric model | Semantic Scholar (2024)

This dissertation aims at extending object class detection and pose estimation tasks on single 2D images by a 3D model-based approach and proposes an alternative probabilistic pose estimation method which evaluates the likelihood of groups of 2D part detections with respect to a full 3D geometry.

This paper addresses view-invariant object detection and pose estimation from a single image, and revisits the viewer-centered framework, and uses image contours as basic features.

This paper revisits ideas from the early days of computer vision, namely, 3D geometric object class representations for recognition, and shows how these representations can recover geometrically far more accurate object hypotheses than just 2D bounding boxes.

This paper revisit ideas from the early days of computer vision, namely, detailed, 3D geometric object class representations for recognition, and demonstrates novel applications enabled by such an object class representation, such as fine-grained categorization of cars and bicycles, according to their 3D geometry, and ultrawide baseline matching.

Experiments on a new car dataset with diverse views show that the proposed method can learn meaningful 3D car templates, and give satisfactory detection and view estimation performance.

A novel approach to recover objects 3D position and occupancy in a generic scene using only 2D object detections from multiple view images is presented and a data preconditioning technique and a non-linear refinement of the closed form solution based on implicit subspace constraints are introduced.

The voting method employs a novel parametrization of joint detection and viewpoint hypothesis space, allowing efficient accumulation of evidence, and combines this with a re-scoring and refinement mechanism, using an ensemble of view-specific Support Vector Machines.

...

...

This approach extracts a set of pose and class discriminant features from synthetic 3D object models using a filtering procedure, evaluates their suitability for matching to real image data and represents them by their appearance and 3D position.

A novel object class detection method based on 3D object modeling that establishes spatial connections between multiple 2D training views by mapping them directly to the surface of 3D model.

This work proposes a new 3D object class model that is capable of recognizing unseen views by pose estimation and synthesis and performs superiorly to and on par with state-of-the-art algorithms on the Savarese et al. 2007 and PASCAL datasets in object detection.

A method for quantifying the degree of difficulty of detecting objects across views and a novel alignment algorithm for pose-based clustering on the view sphere are proposed and an active learning algorithm that searches for local minima of a classifier’s output in a low-dimensional space of rendering parameters is introduced.

An approach to accurately detect and segment partially occluded objects in various viewpoints and scales is introduced and a novel framework for combining object-level descriptions with pixel-level appearance, boundary, and occlusion reasoning is introduced.

This work proposes a novel representation to model 3D object classes that allows the model to synthesize novel views of an object class at recognition time and incorporates it in a novel two-step algorithm that is able to classify objects under arbitrary and/or unseen poses.

A new approach for detecting objects based on measuring the spatial consistency between different parts of an object, where parts are pre-defined on a set of training images and then located in any arbitrary image.

A system that constructs “implicit shape models” for classes of rigid 3D objects and utilizes these models to estimating the pose of class instances in single 2D images and demonstrates the utility of the constructed models by applying them in pose estimation experiments to recover the viewpoint of class instance.

It is found that for object classes that have substantial geometric structure, such as airplanes, faces and motorbikes, a relatively small amount of spatial structure in the model can provide statistically indistinguishable recognition performance from more powerful models, and at a substantially lower computational cost.

...

...