Reconstructing PASCAL VOC

By Sara Alexandra Gomes Vicente, João Carreira, Lourdes de Agapito, Jorge P. Batista. 2014

191 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition, Evaluation & Benchmarks

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Vision and Pattern Recognition

In Evaluation & Benchmarks

Citations in the corpus

Learning Category-Specific Mesh Reconstruction from Image Collections (2018, methodology): “As previously leveraged by [31,15], applying structurefrom-motion to the annotated keypoint locations additionally allows us to obtain a rough”
Multi-view Supervision for Single-View Reconstruction via Differentiable Ray Consistency (2017, methodology): “[35] proposed a nonparametric method by leveraging surrogate instances – but at the cost of requiring annotations at test time.”
3D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstruction (2016, background): “…by early works that used machine learning to learn a 2D-to-3D mapping for scene understanding [29,30], data driven approaches have been recently proposed to solve the daunting problem of recovering the shape of an object from just a single image [31,32] for a given number of object categories.”
Lifting Object Detection Datasets into 3D (2015, methodology): “adriven method for class-based 3D reconstruction that relies only on 2D information, such as ﬁgure-ground segmentations and a few keypoint annotations. This paper extends the original conference work [11] with additional visualizations and many new experiments: we present a direct evaluation of view”
Category-specific object reconstruction from a single image (2014, methodology): “Note that [36] operates with ground truth annotations and reconstructs an image corpus and our method is used here on the same task for a fair comparison.”
Single Image 3D Interpreter Network (2016, methodology): “Depth-/mesh-based models can recover detailed 3D object structure from a single image, either by adapting existing 3D models from a database [3,19,6,8,20,21,22,23,24], or by inferring from its detected 2D silhouette [5,7,4].”
Multi-view 3D Models from Single Images with a Convolutional Network (2015, methodology): “For building their morphable 3D model they rely on ideas from Vicente et al. [19], who showed that the coarse 3D structure can be reconstructed from multiple images of the same object class (but diﬀerent object instances) and some keypoint annotation.”
Sparse Representation for 3D Shape Estimation: A Convex Relaxation Approach (2015, background): “Recently, there was growing interest in reconstructing category-specific object models from a collection of single images of different instances with the same category [19], [20], [21], [22].”
Learning to See by Moving (2015, methodology): “In order to make a holistic evaluation of the utility of features learnt by different pretraining methods on this task, matching error was computed as a function of viewpoint distance [36].”
Learning Category-Specific Deformable 3D Models for Object Reconstruction (2017, methodology): “TABLE 1 Studying the Quality of Our Learnt 3D Models: Comparison between Our Method and [23], [63] Using Ground Truth Keypoints and Masks on PASCAL VOC”
Rethinking Reprojection: Closing the Loop for Pose-Aware Shape Reconstruction from a Single Image (2017, methodology): “In these methods weaker annotations are often required for optimizing shape or pose, such as key points [22] or instance segmentation [9], limiting the application and performance of these methods.”
Pose Induction for Novel Object Categories (2015, methodology): “Ultimately, and the main motivation for this work, it provides important cues about the 3D shape of a novel object and may allow bypassing the existing need for ground truth keypoints in training data for state-of-the-art class-specific object reconstruction systems [21, 38] – we will present a proo”
Using Locally Corresponding CAD Models for Dense 3D Reconstructions from a Single Image (2017, methodology): “To our best knowledge, the most related work [8, 16] reconstruct 3D dense models purely from 2D images without any access to 3D CAD models.”
Unsupervised Learning of 3D Object Categories from Videos in the Wild (2021, methodology): “Early reconstruction methods for 3D object categories used Non-Rigid SfM (NR-SfM) applied to 2D keypoint annotations [4, 58, 3].”
Learning Shape Abstractions by Assembling Volumetric Primitives (2016, unknown): “Apart from these attempts, most mainstream methods for representing and reconstructing objects typically use much higher-dimensional representations e.g. objects as point clouds [19, 35] or exemplar CAD models [23, 24, 38].”
FlowWeb: Joint image set alignment by weaving consistent, pixel-wise correspondences (2015, background): “[39] on reconstructing PASCAL VOC classes using hand-annotated keypoints is an exciting step in this direction.”
A Simple, Fast and Highly-Accurate Algorithm to Recover 3D Shape from 2D Landmarks on a Single Image (2016, background): “Directly related to our work are several approaches that reconstruct 3D shape from a single image [2], [3], [4], [5], [12].”
C3DPO: Canonical 3D Pose Networks for Non-Rigid Structure From Motion (2019, unknown): “[37, 7] reconstruct the categories of PASCAL VOC.”
WarpNet: Weakly Supervised Matching for Single-View Reconstruction (2016, methodology): “Recently, methods have been proposed to match instances across categories, relying on supervision in the form of part (keypoint) annotations [7, 15, 29] or 3D CAD models [3, 8] to augment appearance information with shape priors.”
Image-Based 3D Object Reconstruction: State-of-the-Art and Trends in the Deep Learning Era (2019, background): “Recent techniques showed that, with only 2D annotations, it is possible to build category-speciﬁc 3D morphable models from 2D silhouettes or 2D images [69], [70].”