Tootfinder

ViSTA: Visual Storytelling using Multi-modal Adapters for Text-to-Image Diffusion Models
Sibo Dong, Ismail Shaheen, Maggie Shen, Rupayan Mallick, Sarah Adel Bargal
https://arxiv.org/abs/2506.12198

ViSTA: Visual Storytelling using Multi-modal Adapters for Text-to-Image Diffusion Models
Text-to-image diffusion models have achieved remarkable success, yet generating coherent image sequences for visual storytelling remains challenging. A key challenge is effectively leveraging all previous text-image pairs, referred to as history text-image pairs, which provide contextual information for maintaining consistency across frames. Existing auto-regressive methods condition on all past image-text pairs but require extensive training, while training-free subject-specific approaches ens…

Wieder ein spannendes Wochenende in #chemnitz!
Starker Tipp
https://chaos.social/@chch/114697409182100038

Chaostreff Chemnitz (@chch@chaos.social)
Attached: 1 image Kommt am Wochenende zur Makers United in die Chemnitzer Stadthalle. Freut euch auf unseren Arcade-Automaten, auf @ChaosErika, die wieder fleißig eure Fragen beantwortet (ja, sie tippt noch!), und auf unsere Lötstationenfür Jung und Alt. Außerdem sind zwei Maker aus Polen und den Niederlanden am Start. Wir freuen uns auf eure Fragen, Ideen und Gespräche über Technik, Kultur, Digitalisierung und alles dazwischen und außerhalb.

Mechanism and Stability of Li-Dynamics in Amorphous Li-Ti-P-S Based Mixed Ionic-Electronic Conductor
Selva Chandrasekaran Selvaraj, Daiwei Wang, Donghai Wang, Anh T. Ngo
https://arxiv.org/abs/2506.11199

Mechanism and Stability of Li-Dynamics in Amorphous Li-Ti-P-S Based Mixed Ionic-Electronic Conductor
Mixed ionic-electronic conductor (MIEC) exhibit both high ionic and electronic conductivity to improve the battery performance. In this work, we investigate the mechanism and stability of transport channels in recently developed our MIEC material, amorphous Ti-doped lithium phosphorus sulfide (LPS), using molecular dynamics (MD) simulations with a 99\% accurate machine-learning force field (MLFF) trained on \textit{ab-initio} MD data. The achieved MLFF helps efficient large-scale MD simulations…

Your values are euros and cents. https://ec.social-network.europa.eu/@EUCommission/114699001106153315

Structural Similarity-Inspired Unfolding for Lightweight Image Super-Resolution
Zhangkai Ni, Yang Zhang, Wenhan Yang, Hanli Wang, Shiqi Wang, Sam Kwong
https://arxiv.org/abs/2506.11823

Structural Similarity-Inspired Unfolding for Lightweight Image Super-Resolution
Major efforts in data-driven image super-resolution (SR) primarily focus on expanding the receptive field of the model to better capture contextual information. However, these methods are typically implemented by stacking deeper networks or leveraging transformer-based attention mechanisms, which consequently increases model complexity. In contrast, model-driven methods based on the unfolding paradigm show promise in improving performance while effectively maintaining model compactness through …

MS-UMamba: An Improved Vision Mamba Unet for Fetal Abdominal Medical Image Segmentation
Caixu Xu, Junming Wei, Huizhen Chen, Pengchen Liang, Bocheng Liang, Ying Tan, Xintong Wei
https://arxiv.org/abs/2506.12441

MS-UMamba: An Improved Vision Mamba Unet for Fetal Abdominal Medical Image Segmentation
Recently, Mamba-based methods have become popular in medical image segmentation due to their lightweight design and long-range dependency modeling capabilities. However, current segmentation methods frequently encounter challenges in fetal ultrasound images, such as enclosed anatomical structures, blurred boundaries, and small anatomical structures. To address the need for balancing local feature extraction and global context modeling, we propose MS-UMamba, a novel hybrid convolutional-mamba mo…

CLIP the Landscape: Automated Tagging of Crowdsourced Landscape Images
Ilya Ilyankou, Natchapon Jongwiriyanurak, Tao Cheng, James Haworth
https://arxiv.org/abs/2506.12214

CLIP the Landscape: Automated Tagging of Crowdsourced Landscape Images
We present a CLIP-based, multi-modal, multi-label classifier for predicting geographical context tags from landscape photos in the Geograph dataset--a crowdsourced image archive spanning the British Isles, including remote regions lacking POIs and street-level imagery. Our approach addresses a Kaggle competition\footnote{https://www.kaggle.com/competitions/predict-geographic-context-from-landscape-photos} task based on a subset of Geograph's 8M images, with strict evaluation: exact match accura…

edit: yes i already know what's up with this image yall can stop explaining in increasingly shitty condescending unhelpful ways

First Positronium Lifetime Imaging with Scandium-44 on a Long Axial Field-of-view PET/CT
Lorenzo Mercolli, William M. Steinberger, Pascal V. Grundler, Anzhelika Moiseeva, Saverio Braccini, Maurizio Conti, Pawe{\l} Moskal, Narendra Rathod, Axel Rominger, Hasan Sari, Roger Schibli, Robert Seifert, Kuangyu Shi, Ewa {\L}. St\k{e}pie\'n, Nicholas P. van der Meulen

First Positronium Lifetime Imaging with Scandium-44 on a Long Axial Field-of-view PET/CT
Purpose: 44Sc has been successfully produced, synthesized, labeled and first-in-human studies were conducted some years ago. The decay properties of 44Sc, together with being close to a clinical implementation, make it an ideal candidate for in vivo positronium lifetime measurements. In this study, we investigate the count statistics for ortho-positronium (oPs) measurements with 44Sc. Method: A NEMA image quality phantom was filled with 41.7 MBq of 44Sc dissolved in water and scanned on a com…

CoMemo: LVLMs Need Image Context with Image Memory
Shi Liu, Weijie Su, Xizhou Zhu, Wenhai Wang, Jifeng Dai
https://arxiv.org/abs/2506.06279 https://…

CoMemo: LVLMs Need Image Context with Image Memory
Recent advancements in Large Vision-Language Models built upon Large Language Models have established aligning visual features with LLM representations as the dominant paradigm. However, inherited LLM architectural designs introduce suboptimal characteristics for multimodal processing. First, LVLMs exhibit a bimodal distribution in attention allocation, leading to the progressive neglect of middle visual content as context expands. Second, conventional positional encoding schemes fail to preser…