Explore the latest research from Facebook

In this paper, we explore methods that enable supervised speech enhancement systems to train on real-world degraded speech data. Specifically, we propose a semi-supervised approach for speech enhancement in which we first train a modified vector-quantized variational autoencoder that solves a source separation task.

In this paper we propose to address policy compliance detection via decomposing it into question answering, where questions check whether the conditions stated in the policy apply to the scenario, and an expression tree combines the answers to obtain the label. Despite the initial upfront annotation cost, we demonstrate that this approach results in better accuracy, especially in the cross-policy setup where the policies during testing are unseen in training.

In this work, we evaluate several model compression techniques for QE and find that, despite their popularity in other NLP tasks, they lead to poor performance in this regression setting.

Here, we use a Kleene algebra with diamond operators and countable joins of tests, which embeds IL, and which also is complete for reasoning about the image of the embedding.

This work outlines how we prioritize original news, a critical indicator of news quality. By examining the landscape and lifecycle of news posts on our social media platform, we identify challenges of building and deploying an originality score.

This position paper names and distinguishes these paradigms. Despite substantial overlap, subtle but significant distinctions exert an outsize influence on research. While one evaluation paradigm values creating more intelligent QA systems, the other paradigm values building QA systems that appeal to users.

In video transmission applications, video signals are transmitted over lossy channels, resulting in low-quality received signals. To restore videos on recipient edge devices in real-time, we introduce an efficient video restoration network, EVRNet.

This paper addresses both of these observations. We give a hybrid algorithm that integrates lookahead into the state-based representation of an SMT solver and show that in the vast majority of cases it is possible to compute full lookahead up to depth four on inexpensive theories.

In this paper, we tackle the egopose estimation from a more natural human vision span, where camera wearer can be seen in the peripheral view and depending on the head pose the wearer may become invisible or has a limited partial view.

In this paper, we question if self-supervised learning provides new properties to Vision Transformer (ViT) [16] that stand out compared to convolutional networks (convnets).