Research Area
Year Published

634 Results

June 20, 2018

HHVM JIT: A Profile-Guided, Region-Based Compiler for PHP and Hack

Programming Language Design and Implementation (PLDI)

This paper describes the design of the second generation of the HHVM JIT and how it addresses the challenges to efficiently execute PHP and Hack programs. This new design uses profiling to build an aggressive region-based JIT compiler. We discuss the benefits of this approach compared to the more popular method-based and trace-based approaches to compile dynamic languages.

By: Guilherme Ottoni

June 19, 2018

Link and code: Fast indexing with graphs and compact regression codes

Computer Vision and Pattern Recognition (CVPR)

Similarity search approaches based on graph walks have recently attained outstanding speed-accuracy trade-offs, taking aside the memory requirements. In this paper, we revisit these approaches by considering, additionally, the memory constraint required to index billions of images on a single server.

By: Matthijs Douze, Alexandre Sablayrolles, Hervé Jégou

June 19, 2018

A Generative Adversarial Approach for Zero-Shot Learning from Noisy Texts

Computer Vision and Pattern Recognition (CVPR)

Most existing zero-shot learning methods consider the problem as a visual semantic embedding one. Given the demonstrated capability of Generative Adversarial Networks (GANs) to generate images, we instead leverage GANs to imagine unseen categories from text descriptions and hence recognize novel classes with no examples being seen.

By: Yizhe Zhu, Mohamed Elhoseiny, Bingchen Liu, Xi Peng, Ahmed Elgammal

June 19, 2018

LAMV: Learning to align and match videos with kernelized temporal layers

Computer Vision and Pattern Recognition (CVPR)

This paper considers a learnable approach for comparing and aligning videos. Our architecture builds upon and revisits temporal match kernels within neural networks: we propose a new temporal layer that finds temporal alignments by maximizing the scores between two sequences of vectors, according to a time-sensitive similarity metric parametrized in the Fourier domain.

By: Lorenzo Baraldi, Matthijs Douze, Rita Cucchiara, Hervé Jégou

June 18, 2018

Learning to Segment Every Thing

Computer Vision and Pattern Recognition (CVPR)

The goal of this paper is to propose a new partially supervised training paradigm, together with a novel weight transfer function, that enables training instance segmentation models on a large set of categories all of which have box annotations, but only a small fraction of which have mask annotations.

By: Ronghang Hu, Piotr Dollar, Kaiming He, Trevor Darrell, Ross Girshick

June 18, 2018

Stacked Latent Attention for Multimodal Reasoning

Computer Vision and Pattern Recognition (CVPR)

Attention has shown to be a pivotal development in deep learning and has been used for a multitude of multimodal learning tasks such as visual question answering and image captioning. In this work, we pinpoint the potential limitations to the design of a traditional attention model.

By: Haoqi Fan, Jiatong Zhou

June 18, 2018

3D Semantic Segmentation with Submanifold Sparse Convolutional Networks

Computer Vision and Pattern Recognition (CVPR)

We introduce new sparse convolutional operations that are designed to process spatially-sparse data more efficiently, and use them to develop spatially-sparse convolutional networks.

By: Benjamin Graham, Laurens van der Maaten, Martin Engelcke

June 18, 2018

Deep Spatio-Temporal Random Fields for Efficient Video Segmentation

Computer Vision and Pattern Recognition (CVPR)

In this work we introduce a time- and memory-efficient method for structured prediction that couples neuron decisions across both space at time. We show that we are able to perform exact and efficient inference on a densely connected spatio-temporal graph by capitalizing on recent advances on deep Gaussian random fields.

By: Siddhartha Chandra, Camille Couprie, Iasonas Kokkinos

June 18, 2018

Modeling Facial Geometry using Compositional VAEs

Computer Vision and Pattern Recognition (CVPR)

We propose a method for learning non-linear face geometry representations using deep generative models. Our model is a variational autoencoder with multiple levels of hidden variables where lower layers capture global geometry and higher ones encode more local deformations.

By: Timur Bagautdinov, Chenglei Wu, Jason Saragih, Pascal Fua, Yaser Sheikh

June 18, 2018

CondenseNet: An Efficient DenseNet using Learned Group Convolutions

Computer Vision and Pattern Recognition (CVPR)

In this paper we develop CondenseNet, a novel network architecture with unprecedented efficiency. It combines dense connectivity with a novel module called learned group convolution. 

By: Gao Huang, Shichen Liu, Laurens van der Maaten, Kilian Q. Weinberger