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 “"text alignment"”.
-
Text-image Restoration And Text Alignment For Multi-engine Optical Character Recognition Systems
… image preprocessing and morphological text restoration that goes to OCR engines. This method is similar to the one used in restoration partial finger prints. Series of morphological dilating and eroding filters of various mask shapes and sizes were applied to text of different font …
-
Designing a General Framework for Text Alignment: Case Studies with Two South Asian Languages
… parallel data to learn accurate word alignment. Additionally, these languages are morphologically rich and have free word order. When it is difficult to rely purely on statistical methods due to insufficient data, research shows that better performance can be obtained by building …
-
Ilgų tekstų sugretinimas naudojant semantinius žodynus /
Long text alignment using semantic dictionarys.
-
Audio-Visual Speech Processing for Multimedia Localisation
… A key aspect of the localisation process is the alignment of content, such as subtitles or audio, when adapting content from one region to another. One method of automating this is through using audio content as a guide, providing a solution via audio-to-text alignment. While many approaches for …
-
Seeding alignment between language technology and indigenous methodologies : the making of Kʷu sqilxʷ an interactive digital edition
… edition to date to use automatic speech-to-text alignment for nsyilxcən. The written portion of this Digital Humanities (DH) project addresses a longstanding schism between western and Indigenous ways of knowing that threaten to further a history of disingenuous settler-colonial research. …
-
Advancing vision-language models for open-vocabulary recognition and generative evaluation
… With VLMs, we can recognize objects using text descriptions, answer questions about images, and generate images from text prompts. Despite this progress, many challenges remain. This thesis explores key advancements in VLMs, tests their capabilities, and points toward unanswered questions …
-
TOWARDS TEXT-BASED UNDERSTANDING OF MOLECULES VIA LANGUAGE MODELING
… the interaction between molecular structures and textual data via language modeling. Our goal is to augment molecule research with a text-based interface that utilizes the vast scientific knowledge and mathematical reasoning capability of modern LMs. To this end, we have contributed multi-modal …
-
Synthetic longitudinal case notes for human services: a free-text corpus and utility-centered evaluation
… tabular client records only, and clinical text corpora (e.g., hospital discharge summaries) document single medical encounters rather than longitudinal social-service trajectories. We release, to our knowledge, the first free-text synthetic longitudinal case note corpus for human …
-
The Influence of Digital Aesthetic Cues on Consumer Judgments and Behavior Towards Brands
… explores how an understudied design choice (text alignment) can affect consumer perceptions of brands. The findings show that the straight leftmost edge generated by left-justified text (vs. centered text) increases perceptions of visual verticality upon exposure, which automatically and …
-
Towards Trustworthy and Robust Multi-modal Machine Learning
… use of increasingly diverse data types, such as text, speech, images, and videos, for training, perception, and inference. However, as the multi-modal machine learning models evolve, they have faced significant challenges related to biases and unseen data distribution, undermining their …
-
Multi-Modal Learning for Threat Analysis
… of letting models learn visual concepts from raw text instead of from manual annotations. One specific model using this concept is CLIP [1], which has shown state-of-the art performance on general zero-shot image classification tasks. However, few works have explored how competitive CLIP is in …
-
Augmenting Video Retrieval: Leveraging Auxiliary Information to Enhance Retrieval Tasks
… challenge of bridging the gap between video and text modalities.<br/>Traditional retrieval methods often rely on a textual query (e.g., video caption) or a<br/>visual query (e.g., video clip) to retrieve the corresponding relevant item from a gallery in the other modality. However, all these …
-
The syntax and prosody of interrogatives: evidence from varieties spoken in northern Italy
… marking, (ii) prosodic marking and (iii) tune-text-alignment.</p>
-
Novel Methods for Natural Language Modeling and Pretraining
… the segmentation feature, lexical feature, and alignment feature. We present three new techniques that improve language sequence modeling with different language features. 1. Segment-Aware Language Modeling is a novel model architecture leveraging the text segementation feature for text sequence …
-
Efficient and Controllable Adaptation of Foundation Models for Multimodal Medical and Vision Systems
… while preserving its promptable nature. \textbf{AdaptiveSAM} introduces a lightweight combination of bias‑and‑norm tuning with a Text‑Alignment Module to enable text‑based prompting and efficient domain adaptation. \textbf{S‑SAM} extends this idea by learning affine transformations over …
-
Generative AI for Digital Pathology: Advancing Diffusion, Vision-Language, and Safety-Aware Models
… processes to generating synthetic diagnostic text for enhancing vision-language model (VLM) representation and alignment. The work further explores the integration of safety and context awareness into generative frameworks to significantly improve model performance and trustworthiness. Chapter …
-
Essays on Multimodal Machine Learning and Social Media Analytics
… across multiple modalities, such as images and text, or video and audio. Building on existing literature that explores the impact of attributes such as content and influencer characteristics on social media engagement, this dissertation focuses on how to utilize information from unstructured …
-
Ensemble Methods for Historical Machine-Printed Document Recognition
… directly related to the quality of the extracted text. Optical Character Recognition (OCR) has reached a point where well-formatted and clean machine- printed documents are easily recognizable by current commercial OCR products; however, older or degraded machine-printed documents present problems …
-
The enhancement of machine translation for low-density languages using Web-gathered parallel texts.
… only effective with a large corpus of parallel texts - texts in two languages that are translations of one another - which most languages lack. This thesis describes the Babylon project which attempts to alleviate this shortage by supplementing existing parallel texts with texts gathered …
-
Privacy Protection in Conversations
… thesis tackles the problem of privacy leakage in textual conversations and proposes to mitigate the risks of privacy disclosure by detecting and rewriting the risky utterances. Previous research on privacy protection in text has a focus on manipulating the implicit semantic representations in a …
Page 1 of 2