Jiayun Luo

Jiayun Luo, Siddhesh Khandelwal, Leonid Sigal et al.

From image-text pairs, large-scale vision-language models (VLMs) learn to implicitly associate image regions with words, which prove effective for tasks like visual question answering. However, leveraging the learned association for open-vocabulary semantic segmentation remains a challenge. In this paper, we propose a simple, yet extremely effective, training-free technique, Plug-and-Play Open-Vocabulary Semantic Segmentation (PnP-OVSS) for this task. PnP-OVSS leverages a VLM with direct text-to-image cross-attention and an image-text matching loss. To balance between over-segmentation and under-segmentation, we introduce Salience Dropout; by iteratively dropping patches that the model is most attentive to, we are able to better resolve the entire extent of the segmentation mask. PnP-OVSS does not require any neural network training and performs hyperparameter tuning without the need for any segmentation annotations, even for a validation set. PnP-OVSS demonstrates substantial improvements over comparable baselines (+26.2% mIoU on Pascal VOC, +20.5% mIoU on MS COCO, +3.1% mIoU on COCO Stuff and +3.0% mIoU on ADE20K). Our codebase is at https://github.com/letitiabanana/PnP-OVSS.

Jiayun Luo, Boyang Li, Cyril Leung

In the evolution of agriculture to its next stage, Agriculture 5.0, artificial intelligence will play a central role. Controlled-environment agriculture, or CEA, is a special form of urban and suburban agricultural practice that offers numerous economic, environmental, and social benefits, including shorter transportation routes to population centers, reduced environmental impact, and increased productivity. Due to its ability to control environmental factors, CEA couples well with computer vision (CV) in the adoption of real-time monitoring of the plant conditions and autonomous cultivation and harvesting. The objective of this paper is to familiarize CV researchers with agricultural applications and agricultural practitioners with the solutions offered by CV. We identify five major CV applications in CEA, analyze their requirements and motivation, and survey the state of the art as reflected in 68 technical papers using deep learning methods. In addition, we discuss five key subareas of computer vision and how they related to these CEA problems, as well as eleven vision-based CEA datasets. We hope the survey will help researchers quickly gain a bird-eye view of the striving research area and will spark inspiration for new research and development.

Samuel Cahyawijaya, Holy Lovenia, Joel Ruben Antony Moniz et al. · cambridge

Southeast Asia (SEA) is a region of extraordinary linguistic and cultural diversity, yet it remains significantly underrepresented in vision-language (VL) research. This often results in artificial intelligence (AI) models that fail to capture SEA cultural nuances. To fill this gap, we present SEA-VL, an open-source initiative dedicated to developing high-quality, culturally relevant data for SEA languages. By involving contributors from SEA countries, SEA-VL aims to ensure better cultural relevance and diversity, fostering greater inclusivity of underrepresented languages in VL research. Beyond crowdsourcing, our initiative goes one step further in the exploration of the automatic collection of culturally relevant images through crawling and image generation. First, we find that image crawling achieves approximately ~85% cultural relevance while being more cost- and time-efficient than crowdsourcing. Second, despite the substantial progress in generative vision models, synthetic images remain unreliable in accurately reflecting SEA cultures. The generated images often fail to reflect the nuanced traditions and cultural contexts of the region. Collectively, we gather 1.28M SEA culturally-relevant images, more than 50 times larger than other existing datasets. Through SEA-VL, we aim to bridge the representation gap in SEA, fostering the development of more inclusive AI systems that authentically represent diverse cultures across SEA.

Wan-Cyuan Fan, Jiayun Luo, Declan Kutscher et al.

Vision-Language Models (VLMs) have been shown to be blind, often underutilizing their visual inputs even on tasks that require visual reasoning. In this work, we demonstrate that VLMs are selectively blind. They modulate the amount of attention applied to visual inputs based on linguistic framing even when alternative framings demand identical visual reasoning. Using visual attention as a probe, we quantify how framing alters both the amount and distribution of attention over the image. Constrained framings, such as multiple choice and yes/no, induce substantially lower attention to image context compared to open-ended, reduce focus on task-relevant regions, and shift attention towards uninformative tokens. We further demonstrate that this attention misallocation is the principal cause of degraded accuracy and cross-framing inconsistency. Building on this mechanistic insight, we introduce a lightweight prompt-tuning method using learnable tokens that encourages the robust, visually grounded attention patterns observed in open-ended settings, improving visual grounding and improving performance across framings.

Jiayun Luo, Wan-Cyuan Fan, Lyuyang Wang et al.

Large Vision Language Models (LVLMs) have recently emerged as powerful architectures capable of understanding and reasoning over both visual and textual information. These models typically rely on two key components: a Vision Transformer (ViT) and a Large Language Model (LLM). ViT encodes visual content into a sequence of image tokens and serves as the perceptual front-end -- the eyes of the model. In contrast, the LLM interprets these tokens to perform high-level reasoning, generates responses, and functions as the cognitive core -- the brain of the model. However, it remains unclear which visual tokens contribute most significantly to understanding and reasoning, and how effectively these signals are propagated from ViT to the LLM. While most existing works have focused on identifying attention sinks, low-semantic tokens receiving disproportionately high attention, within the LLM, we shift the focus to the vision encoder by identifying a class of high-norm visual tokens from ViT, referred to as ViT attention sinks -- a problem that has been rarely studied but is indeed very important for LVLMs. Our findings show that these ViT sinks encapsulate high-level semantic concepts from images, allowing the LLM to perform more effective understanding and reasoning. Despite their importance, these sink tokens are often overlooked in existing LVLM architectures. To explore their contribution, we present both qualitative and quantitative analyses of the information embedded in these sink tokens. We also propose both training-free and training-based approaches to better leverage how this information is interpreted by the LLM, and to what extent. By explicitly utilizing these tokens, we demonstrate substantial improvements across a range of LVLMs and visual reasoning tasks, highlighting the untapped potential of ViT attention sinks in enhancing visual reasoning.

Jiayun Luo, Mir Rayat Imtiaz Hossain, Boyang Li et al.

Vision-Language Models (VLMs) implicitly learn to associate image regions with words from large-scale training data, demonstrating an emergent capability for grounding concepts without dense annotations[14,18,51]. However, the coarse-grained supervision from image-caption pairs is often insufficient to resolve ambiguities in object-concept correspondence, even with enormous data volume. Rich semantic and syntactic structures within the text modality have been overlooked as sources of supervision. Starting from contrastive architectures (BLIP and ALBEF) that show strong intrinsic grounding abilities, we propose HIerarchically STructured Learning (HIST). HIST enhances spatial vision-language alignment without using additional human annotations, by hierarchically decomposing captions into the constituent Subjects, Phrases, and Composite Phrases, and enforcing entailment relation between a parent and its children in the hierarchy. Specifically, we introduce two novel loss functions: (1) Subject Loss, which aligns image content with the subject of the corresponding phrase, acting as an entailment of standard contrastive/matching losses at the Phrase level; (2) Composition Loss, to balance attention across multiple objects. HIST is general, and can be applied to any VLM for which attention between vision and language can be computed. Compared to baseline VLMs, HIST achieves up to +9.8% improvement in visual grounding and +6.3% in multi-object referring segmentation. Surprisingly, the improved spatial grounding leads to improvements in other downstream VLM tasks: +1.1% in image-text retrieval, and +0.2% in visual question answering.

Weiwei Jiang, Jiayun Luo

Drought is a serious natural disaster that has a long duration and a wide range of influence. To decrease the drought-caused losses, drought prediction is the basis of making the corresponding drought prevention and disaster reduction measures. While this problem has been studied in the literature, it remains unknown whether drought can be precisely predicted or not with machine learning models using weather data. To answer this question, a real-world public dataset is leveraged in this study and different drought levels are predicted using the last 90 days of 18 meteorological indicators as the predictors. In a comprehensive approach, 16 machine learning models and 16 deep learning models are evaluated and compared. The results show no single model can achieve the best performance for all evaluation metrics simultaneously, which indicates the drought prediction problem is still challenging. As benchmarks for further studies, the code and results are publicly available in a Github repository.

Weiwei Jiang, Jiayun Luo

Big data has been used widely in many areas including the transportation industry. Using various data sources, traffic states can be well estimated and further predicted for improving the overall operation efficiency. Combined with this trend, this study presents an up-to-date survey of open data and big data tools used for traffic estimation and prediction. Different data types are categorized and the off-the-shelf tools are introduced. To further promote the use of big data for traffic estimation and prediction tasks, challenges and future directions are given for future studies.

Weiwei Jiang, Jiayun Luo

Traffic forecasting is important for the success of intelligent transportation systems. Deep learning models, including convolution neural networks and recurrent neural networks, have been extensively applied in traffic forecasting problems to model spatial and temporal dependencies. In recent years, to model the graph structures in transportation systems as well as contextual information, graph neural networks have been introduced and have achieved state-of-the-art performance in a series of traffic forecasting problems. In this survey, we review the rapidly growing body of research using different graph neural networks, e.g. graph convolutional and graph attention networks, in various traffic forecasting problems, e.g. road traffic flow and speed forecasting, passenger flow forecasting in urban rail transit systems, and demand forecasting in ride-hailing platforms. We also present a comprehensive list of open data and source resources for each problem and identify future research directions. To the best of our knowledge, this paper is the first comprehensive survey that explores the application of graph neural networks for traffic forecasting problems. We have also created a public GitHub repository where the latest papers, open data, and source resources will be updated.