Related papers: iBOT: Image BERT Pre-Training with Online Tokenize…
Modern multimodal large language models (MLLMs) typically keep the language model fixed and train a visual projector that maps the pixels into a sequence of tokens in its embedding space, so that images can be presented in essentially the…
Masked Language Modeling (MLM) has proven to be an essential component of Vision-Language (VL) pretraining. To implement MLM, the researcher must make two design choices: the masking strategy, which determines which tokens to mask, and the…
Masked Language Modeling (MLM) is widely used to pretrain language models. The standard random masking strategy in MLM causes the pre-trained language models (PLMs) to be biased toward high-frequency tokens. Representation learning of rare…
Scene text recognition (STR) enables computers to recognize and read the text in various real-world scenes. Recent STR models benefit from taking linguistic information in addition to visual cues into consideration. We propose a novel…
In this paper, we explore the possibility of building a unified foundation model that can be adapted to both vision-only and text-only tasks. Starting from BERT and ViT, we design a unified transformer consisting of modality-specific…
Subword tokenization is an essential part of modern large language models (LLMs), yet its specific contributions to training efficiency and model performance remain poorly understood. In this work, we decouple the effects of subword…
The pre-training of masked language models (MLMs) consumes massive computation to achieve good results on downstream NLP tasks, resulting in a large carbon footprint. In the vanilla MLM, the virtual tokens, [MASK]s, act as placeholders and…
Recently, prompt tuning \cite{lester2021power} has gradually become a new paradigm for NLP, which only depends on the representation of the words by freezing the parameters of pre-trained language models (PLMs) to obtain remarkable…
Tokenization is associated with many poorly understood shortcomings in language models (LMs), yet remains an important component for long sequence scaling purposes. This work studies how tokenization impacts model performance by analyzing…
Text Image Machine Translation (TIMT)-the task of translating textual content embedded in images-is critical for applications in accessibility, cross-lingual information access, and real-world document understanding. However, TIMT remains a…
Masked image modeling, an emerging self-supervised pre-training method, has shown impressive success across numerous downstream vision tasks with Vision transformers. Its underlying idea is simple: a portion of the input image is masked out…
This paper presents a comprehensive study on the tokenization techniques employed by state-of-the-art large language models (LLMs) and their implications on the cost and availability of services across different languages, especially low…
LLMs have demonstrated remarkable capabilities in linguistic reasoning and are increasingly adept at vision-language tasks. The integration of image tokens into transformers has enabled direct visual input and output, advancing research…
Large vision and language models learned directly through image-text associations often lack detailed visual substantiation, whereas image segmentation tasks are treated separately from recognition, supervisedly learned without…
Much of natural language processing is focused on leveraging large capacity language models, typically trained over single messages with a task of predicting one or more tokens. However, modeling human language at higher-levels of context…
Deceptive images can be shared in seconds with social networking services, posing substantial risks. Tampering traces, such as boundary artifacts and high-frequency information, have been significantly emphasized by massive networks in the…
Speech tokenization serves as the foundation of speech language model (LM), enabling them to perform various tasks such as spoken language modeling, text-to-speech, speech-to-text, etc. Most speech tokenizers are trained independently of…
Self-supervised learning (SSL) has produced a diverse landscape of vision transformers (ViTs) whose pretrained representations support a wide range of downstream tasks. Towards a better understanding of these models, a body of work has…
To build scalable models for challenging real-world tasks, it is important to learn from diverse, multi-modal data in various forms (e.g., videos, text, and images). Among the existing works, a plethora of them have focused on leveraging…
To make sense of their surroundings, intelligent systems must transform complex sensory inputs to structured codes that are reduced to task-relevant information such as object category. Biological agents achieve this in a largely autonomous…