Skip to search formSkip to main contentSkip to account menu

Unsupervised Feature Learning via Non-parametric Instance Discrimination

This work forms this intuition as a non-parametric classification problem at the instance-level, and uses noise-contrastive estimation to tackle the computational challenges imposed by the large number of instance classes.
View on IEEE (opens in a new tab)

Large-Scale Long-Tailed Recognition in an Open World

An integrated OLTR algorithm is developed that maps an image to a feature space such that visual concepts can easily relate to each other based on a learned metric that respects the closed-world classification while acknowledging the novelty of the open world.
View on IEEE (opens in a new tab)

Multiclass spectral clustering

This work proposes a principled account on multiclass spectral clustering by solving a relaxed continuous optimization problem by eigen-decomposition and clarifying the role of eigenvectors as a generator of all optimal solutions through orthonormal transforms.
View on IEEE (opens in a new tab)

Long-tailed Recognition by Routing Diverse Distribution-Aware Experts

RIDE aims to reduce both the bias and the variance of a long-tailed classifier by RoutIng Diverse Experts (RIDE), which significantly outperforms the state-of-the-art methods by 5% to 7% on all the benchmarks including CIFAR100-LT, ImageNet-LT and iNaturalist.
View PDF on arXiv (opens in a new tab)

Learning Non-Lambertian Object Intrinsics Across ShapeNet Categories

This work focuses on the non-Lambertian object-level intrinsic problem of recovering diffuse albedo, shading, and specular highlights from a single image of an object, and shows that feature learning at the encoder stage is more crucial for developing a universal representation across categories.
View on IEEE (opens in a new tab)

Direct Intrinsics: Learning Albedo-Shading Decomposition by Convolutional Regression

The strategy is to learn a convolutional neural network that directly predicts output albedo and shading channels from an input RGB image patch, which outperforms all prior work, including methods that rely on RGB+Depth input.
View on IEEE (opens in a new tab)

Orthogonal Convolutional Neural Networks

The proposed orthogonal convolution requires no additional parameters and little computational overhead and consistently outperforms the kernel orthogonality alternative on a wide range of tasks such as image classification and inpainting under supervised, semi-supervised and unsupervised settings.
View on IEEE (opens in a new tab)

Segmentation given partial grouping constraints

It is demonstrated not only that it is possible to integrate both image structures and priors in a single grouping process, but also that objects can be segregated from the background without specific object knowledge.
View on IEEE (opens in a new tab)

FlowWeb: Joint image set alignment by weaving consistent, pixel-wise correspondences

This work proposes an algorithm to jointly bring a set of poorly aligned images into pixel-wise correspondence by estimating a FlowWeb representation of the image set by initializing all edges of this complete graph with an off-the-shelf, pairwise flow method.
View on IEEE (opens in a new tab)

Improving Generalization via Scalable Neighborhood Component Analysis

This work uses a deep neural network to learn the visual feature that preserves the neighborhood structure in the semantic space, based on the Neighborhood Component Analysis (NCA) criterion and devise a mechanism to use augmented memory to scale NCA for large datasets and very deep networks.
View PDF on arXiv (opens in a new tab)
...
By clicking accept or continuing to use the site, you agree to the terms outlined in our Privacy Policy (opens in a new tab), Terms of Service (opens in a new tab), and Dataset License (opens in a new tab)