Sparse Mixture-of-Experts models (MoEs) have recently gained popularity ...
Multimodal learning is defined as learning over multiple heterogeneous i...
Self supervision and natural language supervision have emerged as two
ex...
Recent advances in zero-shot image recognition suggest that vision-langu...
We present the Pathways Autoregressive Text-to-Image (Parti) model, whic...
We study the problem of synthesizing immersive 3D indoor scenes from one...
Recent work has shown that either (1) increasing the input length or (2)...
It has been shown that dual encoders trained on one domain often fail to...
Both image-caption pairs and translation pairs provide the means to lear...
Language agnostic and semantic-language information isolation is an emer...
We provide the first exploration of text-to-text transformers (T5) sente...
People navigating in unfamiliar buildings take advantage of myriad visua...
Pre-trained representations are becoming crucial for many NLP and percep...
The output of text-to-image synthesis systems should be coherent, clear,...
This paper presents a novel training method, Conditional Masked Language...
Localized Narratives is a dataset with detailed natural language descrip...
In this paper we explore the effects of negative sampling in dual encode...
Neural models that independently project questions and answers into a sh...
We adapt multilingual BERT to produce language-agnostic sentence embeddi...
ROUGE is the de facto criterion for summarization research. However, its...
Retrieval question answering (ReQA) is the task of retrieving a
sentence...
Image captioning datasets have proven useful for multimodal representati...
Deep neural scoring models have recently been shown to improve ranking
q...
We present LAReQA, a challenging new benchmark for language-agnostic ans...
Named Entity Recognition systems achieve remarkable performance on domai...
Named entity recognition systems perform well on standard datasets compr...
Most existing work on adversarial data generation focuses on English. Fo...
Existing curriculum learning research in neural machine translation (NMT...
Popular QA benchmarks like SQuAD have driven progress on the task of
ide...
We introduce two pre-trained retrieval focused multilingual sentence enc...
We explore using multilingual document embeddings for nearest neighbor m...
Modern NLP systems require high-quality annotated data. In specialized
d...
In this paper, we present an approach to learn multilingual sentence
emb...
Neural language models have been shown to achieve an impressive level of...
Product reviews, in the form of texts dominantly, significantly help
con...
This paper presents an effective approach for parallel corpus mining usi...
We present a corpus of 5,000 richly annotated abstracts of medical artic...
Medical professionals search the published literature by specifying the ...
We present a novel approach to learn representations for sentence-level
...
We present models for encoding sentences into embedding vectors that
spe...
Content-dense news report important factual information about an event i...
We present a robust approach for detecting intrinsic sentence importance...