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 30 for “"Text Embedding"”.

  1. Domain-Specific Text Embedding Models for Information Retrieval

    … like chemistry, often leads to issues such as context limitations, hallucinations, difficulty updating their parametric knowledge, and unclear sources for generated responses. To tackle these challenges, Retrieval-Augmented Generation (RAG) enables language models to use external knowledge sources …

    mcmaster Repository record for Domain-Specific Text Embedding Models for Information Retrieval (opens in a new tab)

  2. Depression detection in social media via differential text embedding

    … this problem by generating \extit{differential embeddings} using the Sentence BERT transformer architecture. More specifically, we proposed a few-shot model that can leverage state-of-the-art (SOTA) representation learning techniques and used it in supervised and unsupervised tasks.We …

    msu Repository record for Depression detection in social media via differential text embedding (opens in a new tab)

  3. Theory-enhanced automation of the digital publics' relationship assessments

    … Processing (NLP) techniques, specifically text-embedding and classification, are used to analyze the crawled data and three survey data. The results demonstrate that TF-IDF, BERT embedding, and the SVM classification model perform best. The case study outcomes using TripAdvisor and …

    oklahoma-thes Repository record for Theory-enhanced automation of the digital publics' relationship assessments (opens in a new tab) Restricted access

  4. Repurposing large pretrained diffusion models for unsupervised visual understanding and efficient adaptation

    Large pretrained text-conditioned image generation models learn a compositional and structured latent representation of visual concepts, showcasing their rich understanding of the world through their ability to generate diverse, coherent images. These models link text descriptions to visual …

    ubc Repository record for Repurposing large pretrained diffusion models for unsupervised visual understanding and efficient adaptation (opens in a new tab) Download PDF for Repurposing large pretrained diffusion models for unsupervised visual understanding and efficient adaptation (opens in a new tab)

  5. Fine-Grained Text-to-Shape Generation via CLIP Latent Space Adaptation

    Generating 3D shapes from text is challenging due to the limited availability of text-to-3D datasets. Existing methods train generative models by conditioning on multi-view CLIP [38] image embeddings of shapes during training and use CLIP text embeddings for inference. However, due to the nature of …

    carleton Repository record for Fine-Grained Text-to-Shape Generation via CLIP Latent Space Adaptation (opens in a new tab)

  6. Explanation and Visualisation Frameworks for Fine-Grained Understanding of Human-Generated Text and Image Data

    … nuanced or human-like understanding of text and image data. We contribute several theoretical and technical findings to explainable AI (XAI) and visual analytics.<br/><br/>Firstly, we propose an explanation framework for fine-grained text classification. The prevalent feature importance …

    kings Repository record for Explanation and Visualisation Frameworks for Fine-Grained Understanding of Human-Generated Text and Image Data (opens in a new tab) Download PDF for Explanation and Visualisation Frameworks for Fine-Grained Understanding of Human-Generated Text and Image Data (opens in a new tab)

  7. On Post-Hoc Dimensionality Reduction of Text Embeddings

    Text embeddings are a core representation in modern NLP, supporting tasks such as retrieval, clustering, classification, and semantic search. However, embeddings often have hundreds or thousands of dimensions, creating substantial storage and efficiency challenges at scale. In this work, we present …

    athens Repository record for On Post-Hoc Dimensionality Reduction of Text Embeddings (opens in a new tab)

  8. Utveckling av ett AI-drivet stödsystem för analys av kundfeedback

    … om produkten, genom att använda AI-modellerna text-embedding-3-large och GPT-4o för datainsamling och textsvar. Testerna visar också att systemet kan besvara både generella och specifika frågor, samt att det har förmåga att resonera kring innehållet i frågorna. Detta tyder på att systemet kan …

    chalmers Repository record for Utveckling av ett AI-drivet stödsystem för analys av kundfeedback (opens in a new tab)

  9. clip-mesh: generating textured meshes from text using pretrained image-text models

    … introduces a novel technique for generating textured mesh models without any 3D supervision based solely on a text prompt. This is done by deforming the control shape of a limit subdivided surface along with its texture and normal map to match an input text prompt. The generated mesh asset …

    concordia Download PDF for clip-mesh: generating textured meshes from text using pretrained image-text models (opens in a new tab)

  10. Semantic Text-to-Face Generative Adversarial Network

    … for generating facial images from semantic text descriptions, which makes progress toward this goal. A model is provided with a text description of facial features and an outline of the face. The model uses these inputs to “sketch” the features. Our models are trained using an Affine …

    buffalo Repository record for Semantic Text-to-Face Generative Adversarial Network (opens in a new tab)

  11. Don’t Judge a Malware by its Binary

    … similarity. The model could create valuable embedding spaces for classifying malware by family. Text embeddings for the reports were computed with both BM25 combined with cosine similarity and a transformer encoder. When we projected sandbox reports into text-embedding space, both BM25 …

    chalmers Repository record for Don’t Judge a Malware by its Binary (opens in a new tab)

  12. Binary Representation Learning for Large Scale Visual Data

    … (CNNs) as backbones, and effective binary embedding algorithm that is seamlessly integrated into the learning process, we achieve state-of-the art performance on several settings. First, we study the supervised binary representation learning problem by using label information directly …

    utk-theses Repository record for Binary Representation Learning for Large Scale Visual Data (opens in a new tab)

  13. Novel Transformer Architectures For 3d Multi-Modal And Multi-Organ Medical Image Segmentation

    … capturing long-range dependencies and global context in 2D CT-based prostate segmentation. FocalUNETR employs focal self-attention mechanisms and incorporates an auxiliary boundary-aware regression task to enhance segmentation precision, particularly in cases with unclear boundaries. Second, we …

    wayne-thes Repository record for Novel Transformer Architectures For 3d Multi-Modal And Multi-Organ Medical Image Segmentation (opens in a new tab)

  14. Multimodal Retrieval Augmented Generation via Embedding Space Alignment

    … relevant information is often spread across text, maps, diagrams, and photographs. Many deployed pipelines are built around strong text embedding models like E5, and fully migrating to a new multimodal retriever can be costly and disruptive. This thesis investigates whether visual retrieval …

    chalmers Repository record for Multimodal Retrieval Augmented Generation via Embedding Space Alignment (opens in a new tab)

  15. Addressing Domain Shifts for Computer Vision Applications via Language

    … DASS framework that enhances spatial context-awareness by leveraging vision-language models (VLMs). LangDA generates scene-level descriptions (e.g., "a pedestrian is on the sidewalk, and the street is lined with buildings") to encode object relationships. At an image-level, LangDA aligns …

    waterloo Repository record for Addressing Domain Shifts for Computer Vision Applications via Language (opens in a new tab)

  16. An image-text model for effective retrieval of seventeenth-century Spanish American notary records

    … thesis, we investigate the potential of an image-text model that captures the semantic similarities between image and text pairs, for effective retrieval of documents in the seventeenth-century Spanish American notary records containing diverse writing styles. We propose a new technique using …

    missouri Repository record for An image-text model for effective retrieval of seventeenth-century Spanish American notary records (opens in a new tab)

  17. The Influence of Structural Information on Natural Language Processing

    … and efficient vectoral representations for text has been a core problem for many downstream tasks in natural language processing (NLP).</p><p>Most traditional NLP approaches learn a text representation by only modeling the text itself.</p><p>Recently, researchers have discovered that some …

    duke Repository record for The Influence of Structural Information on Natural Language Processing (opens in a new tab)

Page 1 of 2