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 28 for “"image-text"”.
-
Ramble: To Wander & Wayfind in Image & Text
… particular, I am interested in the ways in which image and text might ramble (move or talk seemingly without set destination) alongside each other in purposeful ambiguity. As Rebecca Solnit states, it is in the “indeterminacy of a ramble” that there “much is to be discovered.”</p>
-
Hybrid ConVIRT - enhancing medical image-text representation learning of vision language models
… are being addressed through advancements in image-text representation learning, as demonstrated by Hybrid-ConVIRT, which builds on contrastive learning frameworks such as ConVIRT and MedCLIP. These medical contrastive learning models trained on domain-specific datasets, have tackled issues …
-
Image, text and the female body : René Magritte and the surrealist publications
… drawing, "Le Viol" (The Rape) on its cover. The image, a view of a woman's head in which her facial features have been replaced by her torso, was meant to shock the viewer out of complacent acceptance of present reality into "surreality," that liberated state of being which would foster …
-
An image-text model for effective retrieval of seventeenth-century Spanish American notary records
… search due to their physical format (e.g., poor image quality) and varied handwriting. The existing work on handwritten documents has predominantly utilized traditional optical character recognition (OCR) techniques, which struggle with the varied and complex handwriting styles. In this thesis, …
-
“I'm No Cactus Expert, But I Know A Prick When I See One” -Image, Text, Therapy and Vengeance in Art by Women
This dissertation examines the roles of text and image, therapy and vengeance in art made by women since the 20th century. It discusses the uses of text and image to create political artworks that can play important roles in feminism in regards to the artistic world and how often vengeance is used …
-
Visual Language Pretrained Multiple Instance Zero-Shot Transfer for Histopathology Images
… method for either training new language-aware image encoders or augmenting existing pretrained models with zero-shot visual recognition capabilities. However, existing works typically train on large datasets of image-text pairs and have been designed to perform downstream tasks involving only …
-
THINGS WE HAVE IN COMMON: ESSAYS AND EXPERIMENTS
… a collection of short stories, flash pieces, and image-text experiments that attempts, in the wake of the death of my mother, to excavate the relationship between memory and narrative, identity and belonging against a backdrop of the main forces that have influenced my familial group, namely …
-
Moments of Seeing: Woolf, Lewis, and Modernist Exteriority
… For both Woolf and Lewis, the object, the visual image, or what this study calls exteriority, retains the power to "make it new," a phrase always associated with the modernist project, but with an important difference. Pound's imagism emphasizes the strict avoidance of all clichéd expressions and …
-
Data-Driven General Purpose Foundation Models for Computational Pathology
… general-purpose encoder models for pathology images: one using paired image-text data, and another leveraging self-supervised learning on large-scale unlabeled images. Additionally, I will examine downstream applications of these foundation models, including zero-shot transfer to gigapixel …
-
Data curation for foundation model training
… ways to increase the quality of datasets in image-text and language domains, with a focus on dataset curation and filtering. Given the amount of readily available data on the web, these techniques can be reliably applied as methods to increase the downstream performance of models, with the …
-
Communicating CSR and Brand Personality through Social Media
… companies' CSR initiatives through their text and image posts on Instagram and Twitter, as well as how socially responsible companies express brand personality using these social media sites. Furthermore, this study compared organizations' use of image, text, and text-only based social …
-
Images in motion?: a first look into video leakage in federated learning
… While the impact of these attacks is known for image, text, and tabular data, their effect on video data remains an unexamined area of research. This paper presents the first analysis of video data leakage in FL via gradient inversion attacks. We evaluate two common video classification …
-
Picturing the Text: Illustrated Editions of Marlitt, Raabe, and Storm in the Age of the Industrial Book: (1857-90)
… mode for adults. Informed by book history and image-text studies, I argue that book illustration provides insights into how literary works were read and re-interpreted in the second half of the nineteenth century. My study examines six illustrated book editions by three German authors: Wilhelm …
-
Towards a Unified Framework for Visual Recognition and Generation via Masked Generative Modeling
… begin with MAGE, a novel framework that unifies image generation and recognition while achieving state-ofthe-art performance on both tasks. We then extend it into vision-language multi-modal training through ITIT, which utilizes unpaired image and text data to train models capable of …
-
Data-Efficient Machine Learning with Focus on Transfer Learning
… topics: distant domain TL and cross-modality TL (image-text). A detailed algorithm introduction and preliminary results on real-world applications (Covid-19 diagnosis and image classification) will be presented. Then, I will discuss the current trends in TL algorithms and real-world applications. …
-
It's a kind of magic: exploring the role of the living in effecting food and drink offerings in private mortuary cults of late Old Kingdom Egypt
… a permanent, ‘magical’ supply was provided via image, text, and object representations of sustenance in the tomb. This study examines these ‘modes’ through which offerings were present and presented. Complementary theoretical frameworks are employed in case studies that analyse illustrative …
-
Multimodal Representation Learning for Agentic AI Systems
… utilizes progressive self-distillation and soft image-text alignments to model the many-to-many correspondences found in noisy web-harvested datasets. Extensive evaluation demonstrates that our method consistently outperforms CLIP across multiple benchmarks, including improved robustness to …
-
Leveraging Retrieval-Augmented Generation, Prompt Engineering, and Vision Language Models for Surface Defect Classification and Root Cause Analysis in Manufacturing
… developed by the author. Results showed that image-only prompting consistently outperformed multimodal (image + text) inputs, achieving up to 94.4% defect identification and 82.6% classification accuracy on the #3DBenchy dataset, and 90.5% defect identification with 76% classification accuracy …
-
Imagining space and place: the representation of Africa through image and text in Andrew Lang's Fairy Books (1889-1910)
… as both Victorians and women, shaped the texts through their own sensitivities. The images, also created through one pictorial lens by Henry Justice Ford, were informed by imagination rather than fact, and the images were embraced for artistic merit rather than accuracy. The dissertation …
-
Representational alignment of humans and machines for computer vision
… while traditional training methods embed similar images close together, thereby often ignoring the global organization of object concepts. To address this mismatch, we developed a novel method that improves the global structure of these representations by linearly aligning them with human …
Page 1 of 2