2 code implementations • 10 Apr 2024 • Jinghong Chen, Weizhe Lin, Jingbiao Mei, Bill Byrne
The Directed Acyclic Transformer is a fast non-autoregressive (NAR) model that performs well in Neural Machine Translation.
no code implementations • 17 Mar 2024 • Igor Sterner, Weizhe Lin, Jinghong Chen, Bill Byrne
Two approaches have emerged to input images into large language models (LLMs).
1 code implementation • 13 Feb 2024 • Weizhe Lin, Jingbiao Mei, Jinghong Chen, Bill Byrne
Large Multimodal Models (LMMs) excel in natural language and visual understanding but are challenged by exacting tasks such as Knowledge-based Visual Question Answering (KB-VQA) which involve the retrieval of relevant information from document collections to use in shaping answers to questions.
Ranked #1 on Retrieval on InfoSeek (using extra training data)
no code implementations • 14 Nov 2023 • Jingbiao Mei, Jinghong Chen, Weizhe Lin, Bill Byrne, Marcus Tomalin
Hateful memes have emerged as a significant concern on the Internet.
Ranked #2 on Meme Classification on Hateful Memes
1 code implementation • 14 Nov 2023 • Guangyu Yang, Jinghong Chen, Weizhe Lin, Bill Byrne
Minimum Bayes Risk (MBR) decoding can significantly improve translation performance of Multilingual Large Language Models (MLLMs).
1 code implementation • NeurIPS 2023 • Weizhe Lin, Jinghong Chen, Jingbiao Mei, Alexandru Coca, Bill Byrne
FLMR addresses two major limitations in RA-VQA's retriever: (1) the image representations obtained via image-to-text transforms can be incomplete and inaccurate and (2) relevance scores between queries and documents are computed with one-dimensional embeddings, which can be insensitive to finer-grained relevance.
Ranked #1 on Retrieval on OK-VQA
no code implementations • 23 Sep 2023 • Alexandru Coca, Bo-Hsiang Tseng, Jinghong Chen, Weizhe Lin, Weixuan Zhang, Tisha Anders, Bill Byrne
Schema-guided dialogue state trackers can generalise to new domains without further training, yet they are sensitive to the writing style of the schemata.
no code implementations • 19 Mar 2023 • Weizhe Lin, Zhilin Wang, Bill Byrne
The widely used Fact-based Visual Question Answering (FVQA) dataset contains visually-grounded questions that require information retrieval using common sense knowledge graphs to answer.
1 code implementation • 29 Jan 2023 • Jinghong Chen, Weizhe Lin, Bill Byrne
We show that SGSAcc can be applied to evaluate utterances generated from a wide range of dialogue actions in the Schema Guided Dialogue (SGD) dataset with good agreement with human judgment.
1 code implementation • 7 Oct 2022 • Weizhe Lin, Bill Byrne
The strong retrieval ability of our model significantly reduces the number of retrieved documents needed in training, yielding significant benefits in answer quality and computation required for training.
Ranked #2 on Retrieval on OK-VQA
no code implementations • 2 Apr 2022 • Weizhe Lin, Linjun Shou, Ming Gong, Pei Jian, Zhilin Wang, Bill Byrne, Daxin Jiang
Knowledge graph (KG) based Collaborative Filtering is an effective approach to personalizing recommendation systems for relatively static domains such as movies and books, by leveraging structured information from KG to enrich both item and user representations.
1 code implementation • EMNLP 2021 • Weizhe Lin, Bo-Hsiang Tseng, Bill Byrne
Dialogue State Tracking is central to multi-domain task-oriented dialogue systems, responsible for extracting information from user utterances.
Ranked #1 on Multi-domain Dialogue State Tracking on MULTIWOZ 2.0
1 code implementation • 26 Nov 2020 • QingBiao Li, Weizhe Lin, Zhe Liu, Amanda Prorok
Our Message-Aware Graph Attention neTwork (MAGAT) is based on a key-query-like mechanism that determines the relative importance of features in the messages received from various neighboring robots.
no code implementations • NAACL (NUSE) 2021 • Zhilin Wang, Weizhe Lin, Xiaodong Wu
While many different aspects of human experiences have been studied by the NLP community, none has captured its full richness.
1 code implementation • 31 Jul 2020 • Weizhe Lin, Indigo Orton, QingBiao Li, Gabriela Pavarini, Marwa Mahmoud
Compared to modalities such as face, head, and vocal, research investigating the use of the body modality for these tasks is relatively sparse.
Ranked #1 on Anxiety Detection on Well-being Dataset
no code implementations • 17 Mar 2020 • Xiaodong Wu, Weizhe Lin, Zhilin Wang, Elena Rastorgueva
Online forums and social media platforms provide noisy but valuable data every day.
no code implementations • WS 2019 • Zhilin Wang, Elena Rastorgueva, Weizhe Lin, Xiaodong Wu
This model is built upon the BERT Next Sentence Prediction model and reduces the time complexity for clustering all posts in a corpus from O(n{\^{}}2) to O(n) with respect to the number of posts.