6 citations · 18 across the 14 of their papers we have counts for
7 papers · 2 filters
Advancing Image Retrieval with Few-Shot Learning and Relevance Feedback
Boaz Lerner, Nir Darshan, Rami Ben-Ari
With such a massive growth in the number of images stored, efficient search in a database has become a crucial endeavor managed by image retrieval systems. Image Retrieval with Rel…
Lightning-Fast Image Inversion and Editing for Text-to-Image Diffusion Models
Dvir Samuel, Barak Meiri, Haggai Maron +5
Diffusion inversion is the problem of taking an image and a text prompt that describes it and finding a noise latent that would generate the exact same image. Most current determin…
Watch Where You Head: A View-biased Domain Gap in Gait Recognition and Unsupervised Adaptation
Gavriel Habib, Noa Barzilay, Or Shimshi +2
Gait Recognition is a computer vision task aiming to identify people by their walking patterns. Although existing methods often show high performance on specific datasets, they lac…
Norm-guided latent space exploration for text-to-image generation
Dvir Samuel, Rami Ben-Ari, Nir Darshan +2
Text-to-image diffusion models show great potential in synthesizing a large variety of concepts in new compositions and scenarios. However, the latent space of initial seeds is sti…
Chatting Makes Perfect: Chat-based Image Retrieval
Matan Levy, Rami Ben-Ari, Nir Darshan +1
Chats emerge as an effective user-friendly approach for information retrieval, and are successfully employed in many domains, such as customer service, healthcare, and finance. How…
Generating images of rare concepts using pre-trained diffusion models
Dvir Samuel, Rami Ben-Ari, Simon Raviv +2
Text-to-image diffusion models can synthesize high-quality images, but they have various limitations. Here we highlight a common failure mode of these models, namely, generating un…
Data Roaming and Quality Assessment for Composed Image Retrieval
Matan Levy, Rami Ben-Ari, Nir Darshan +1
The task of Composed Image Retrieval (CoIR) involves queries that combine image and text modalities, allowing users to express their intent more effectively. However, current CoIR…