Global ETD Search

Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.

Results

Showing 1 to 20 of 26 for “"Visual grounding"”.

  1. Visual grounding through iterative refinement

    The problem of visual grounding has attracted much attention in recent years due to its pivotal role in more general visio-lingual high level reasoning tasks (e.g., image captioning, VQA). Despite the tremendous progress in this area, the performance of most approaches has been hindered by the …

    ubc Repository record for Visual grounding through iterative refinement (opens in a new tab) Download PDF for Visual grounding through iterative refinement (opens in a new tab)

  2. Optimizing visual grounding of latent representations of speech from distant language groups

    … research interest into using multi-modal grounding techniques to bolster classic natural language processing (NLP) and automated speech recognition (ASR) tasks. Previous work by Harwath et al. [5], demonstrated that visual grounding approximately doubled their model's bilingual utterance …

    texas Repository record for Optimizing visual grounding of latent representations of speech from distant language groups (opens in a new tab)

  3. Vision and Language: Information Integration and Transformation

    … multi-modalities, such as image-text retrieval, visual question answering, visual grounding, and image captioning. In particular, we explore and develop techniques combining image and text information to improve transformation between modalities. First, we exploit the multilingual image …

    rice Repository record for Vision and Language: Information Integration and Transformation (opens in a new tab)

  4. Towards Transparent and Grounded Visual AI Systems

    … research tries to answer the question -- Do deep visual models make their decisions for the "right reasons"? In my dissertation, I try to answer this question in two ways: 1. Visual grounding. Grounding is essential to build reliable and generalizable systems that are not driven by dataset biases. …

    gatech Repository record for Towards Transparent and Grounded Visual AI Systems (opens in a new tab)

  5. Approaches for Interactions in Robotics Applications

    … trajectories. In the second part, we propose visual grounding models in 3D pointcloud and RGBD images as essential tasks for robot navigation and human-robot interaction. The task is to identify the referred object given language description either from a reconstructed 3D scene or a pair of …

    washington Repository record for Approaches for Interactions in Robotics Applications (opens in a new tab)

  6. Describing and retrieving visual content using natural language

    … methods have boosted research progress in visual recognition and text understanding but it is a non-trivial task to unite these advances from both disciplines. In this thesis, we develop models and techniques that allow us to connect natural language and visual content enabling automatic …

    bu Repository record for Describing and retrieving visual content using natural language (opens in a new tab)

  7. ADVANCING VISION-LANGUAGE PRETRAINING: UNIFIED ARCHITECTURES AND IMPROVED GROUNDING ABILITIES

    The fusion of visual and language pre-training models has significantly advanced AI, particularly in content retrieval, visual reasoning, and multimodal understanding. However, challenges remain in achieving fine-grained semantic alignment and efficient processing of temporal dynamics in video. …

    nus Repository record for ADVANCING VISION-LANGUAGE PRETRAINING: UNIFIED ARCHITECTURES AND IMPROVED GROUNDING ABILITIES (opens in a new tab) Download file for ADVANCING VISION-LANGUAGE PRETRAINING: UNIFIED ARCHITECTURES AND IMPROVED GROUNDING ABILITIES (opens in a new tab)

  8. Not All Policies Are Equal: Training Flow-Based Policies for Efficient Downstream Steering

    … state and observation-sensitivity penalties on visual and proprioceptive conditioning. In custom flow policies, these regularizers improve latent conditioning, produce broader but structured action coverage, and enable successful downstream steering in real-world manipulation tasks. In contrast, …

    washington Repository record for Not All Policies Are Equal: Training Flow-Based Policies for Efficient Downstream Steering (opens in a new tab)

  9. Sub-Question Oriented Tuning of Visual Question Answering Models for Improved Consistency

    Recent research in Visual Question Answering (VQA) has revealed state-of-the-art models to be inconsistent in their understanding of the world -- they answer seemingly difficult questions requiring reasoning correctly but get simpler associated sub-questions wrong. These sub-questions pertain to …

    gatech Repository record for Sub-Question Oriented Tuning of Visual Question Answering Models for Improved Consistency (opens in a new tab)

  10. Large-Scale Offline Pre-Training Bootstraps Embodied Intelligence

    … intelligence. In part 1, we present VC-1, a visual foundation model pre-trained (primarily) on video data collected from an egocentric perspective. We systematically demonstrate that such models substantially benefit from pre-training dataset diversity by introducing CortexBench, an embodied …

    gatech Repository record for Large-Scale Offline Pre-Training Bootstraps Embodied Intelligence (opens in a new tab)

  11. Design as inquiry: A classroom-embedded investigation of aesthetic framing and children’s interpretations of assistive devices in early childhood classrooms

    … involved flashcards depicting assistive then in visually expressive redesigned versions. The activity was conducted in two classrooms by educators, while observational data were collected externally and supplemented by open-ended educator questionnaires. Findings suggest that children interpreted …

    iastate Repository record for Design as inquiry: A classroom-embedded investigation of aesthetic framing and children’s interpretations of assistive devices in early childhood classrooms (opens in a new tab)

  12. A Vision-Language Model Framework for Geographically And Environmentally Informed Soundscape Assessment

    … and demonstrating cross-modal alignment between visual scenes and acoustic properties. To ensure interpretability, I employed minimalist structured prompts requiring explicit visual grounding (e.g., "fountain in center"), supplemented by Grad-CAM and attention rollout visualizations that confirm …

    south-carolina Repository record for A Vision-Language Model Framework for Geographically And Environmentally Informed Soundscape Assessment (opens in a new tab)

  13. Visual understanding of the physical world

    … properties of the scene, and (3) bridging visual understanding with language. For all the themes, we build our method on top of large-scale pre-trained models or their representations.<br><br>For occlusion understanding and handling, we design a tri-layer plugin for conventional pre-trained …

    oxford Repository record for Visual understanding of the physical world (opens in a new tab)

  14. Construction safety inspection with large pre-trained vision-language models

    … posters collected from the internet (i.e., visual inputs) with detailed descriptions of the poster written manually by the researchers (i.e., natural language inputs). Moreover, there is a lack of open datasets to comprehensively evaluate and further fine-tune VLMs in construction safety …

    ubc Repository record for Construction safety inspection with large pre-trained vision-language models (opens in a new tab) Download PDF for Construction safety inspection with large pre-trained vision-language models (opens in a new tab)

  15. Enhancing Computer Vision with Better Data and Inference

    … that enables test-time self-correction for visual grounding tasks, allowing models to refine their predictions through iterative feedback. Together, these contributions advance the robustness, adaptability, and alignment of modern vision and multimodal AI systems. More broadly, this work …

    toronto-retro Repository record for Enhancing Computer Vision with Better Data and Inference (opens in a new tab)

  16. Exploration of Visual, Acoustic, and Physiological Modalities to Complement Linguistic Representations for Sentiment Analysis

    … linguistic representations with acoustic, visual, and physiological features. The potential benefits of using these modalities include linguistic disambiguation, visual grounding, and the integration of information about people's internal states. The main goal of this work is to build …

    unt Repository record for Exploration of Visual, Acoustic, and Physiological Modalities to Complement Linguistic Representations for Sentiment Analysis (opens in a new tab)

  17. Beyond frames: multimodal video understanding

    … of the world is inherently multimodal, combining visual, auditory, textual, and other sensory inputs to create a comprehensive perspective. Similarly, computer vision methods should go beyond visual data alone, incorpo- rating multiple modalities such as motion cues, audio, textual annotations, …

    freiburg-diss Repository record for Beyond frames: multimodal video understanding (opens in a new tab)

  18. Reliable Medical LLM and Vision-Language RAG through Multi-Agent Orchestration and Single-Step Preference Alignment

    … factual accuracy while integrating complex visual and textual information. Our innovative approach addresses these challenges through a unified Triple Preference Optimization framework with three-stage preference dataset curation, focusing on cross-modal alignment, retrieval balance, and a …

    rice Repository record for Reliable Medical LLM and Vision-Language RAG through Multi-Agent Orchestration and Single-Step Preference Alignment (opens in a new tab) Embargoed until 2030-12-01

  19. Learning joint latent representations for images and language

    … labels to generating rich descriptions of visual data, in particular, in the form of natural language. Learning the joint latent representations for images and language is vital to solving many image-text tasks, including image-sentence retrieval, visual grounding, and image captioning, …

    uiuc Repository record for Learning joint latent representations for images and language (opens in a new tab)

  20. Towards Generalizable Representation Learning Across Domains

    … by estimating uncertainty, strengthening visual grounding, and refining model predictions to reduce language bias and hallucination. For deepfake detection, this thesis proposes FreqDebias, a framework designed to improve generalization to unseen forgery types. Many deepfake detectors rely …

    clemson Repository record for Towards Generalizable Representation Learning Across Domains (opens in a new tab)

Page 1 of 2