Tootfinder

@leftsidestory@mstdn.social
2025-10-29 00:35:19

Moody Urbanity - Literally 📄
情绪化城市 - 字面上 📄
📷 Pentax MX
🎞️ Ilford HP5 Plus 400，expired 1993
#filmphotography #Photography #blackandwhite

Ilford HP5 Plus 400 (FF)

English Alt Text:
A black-and-white image showing a person walking across a tiled pavement. Only the lower legs and feet are visible; the person wears dark pants and white shoes. The pavement features geometric patterns, including arrows and diamond shapes, embedded into the tiles. The image has a grainy texture, giving it an artistic, vintage appearance. The scene captures motion and design in an urban setting.

中文替代文本：
一张黑白照片，展示一个人走在铺有几何图案的地砖上。画面中只看到其腿部和脚部，穿着深色裤子和白色鞋…

Ilford HP5 Plus 400 (FF)

English Alt Text:
A close-up view of a large metal pipe, likely part of a water or gas pipeline system. The pipe is joined by a flange with several bolts and has a handwritten label attached. The label reads “招租不做饭者优先 1361302812”. The background shows a landscape with rows of trees, suggesting the pipe is part of a infrastructure. The scene is outdoors, with natural light and a clear sky.

中文替代文本：
这是一张大型金属管道的特写图像，可能属于供水或天然气管道系统。管道通过法兰和多个螺栓连接，顶部贴有手写标签，内容为“招租不做饭者优先 136…

Ilford HP5 Plus 400 (FF)

English Alt Text:
Three safety signs mounted on a tiled wall. Each sign features a red prohibition symbol and bilingual text in Chinese and English. From left to right: “No Burning” with a flame and matchstick icon; “No Entry” with a person walking through a doorway; and “No Climbing” with a person climbing a ladder. The signs are clear and bold, designed to prevent fire hazards, unauthorized access, and unsafe climbing in a restricted area.

中文替代文本：
三块安全警示牌安装在瓷砖墙上。每块牌…

Ilford HP5 Plus 400 (FF)

English Alt Text:
A storefront entrance with glass doors and signage above. The sign includes Chinese text and an image of a bowl, suggesting a food establishment. The text reads “随便外卖外卖盲盒吃出惊喜,” promoting takeout meals with a surprise element. A unit number “0412b” is visible above the door. The setting is urban, with reflections of buildings in the glass. The image is in black and white, giving it a stylized, vintage feel.

中文替代文本：
一家餐饮店的入口，设有玻璃门和上方招牌。招牌上有中文文字和一个碗的图…

@arXiv_csCV_bot@mastoxiv.page
2025-10-01 11:53:27

Stitch: Training-Free Position Control in Multimodal Diffusion Transformers
Jessica Bader, Mateusz Pach, Maria A. Bravo, Serge Belongie, Zeynep Akata
https://arxiv.org/abs/2509.26644

Stitch: Training-Free Position Control in Multimodal Diffusion Transformers
Text-to-Image (T2I) generation models have advanced rapidly in recent years, but accurately capturing spatial relationships like "above" or "to the right of" poses a persistent challenge. Earlier methods improved spatial relationship following with external position control. However, as architectures evolved to enhance image quality, these techniques became incompatible with modern models. We propose Stitch, a training-free method for incorporating external position control into Multi-Modal Dif…

@arXiv_csLG_bot@mastoxiv.page
2025-10-01 11:57:47

Linking Process to Outcome: Conditional Reward Modeling for LLM Reasoning
Zheng Zhang, Ziwei Shan, Kaitao Song, Yexin Li, Kan Ren
https://arxiv.org/abs/2509.26578 https://

Linking Process to Outcome: Conditional Reward Modeling for LLM Reasoning
Process Reward Models (PRMs) have emerged as a promising approach to enhance the reasoning capabilities of large language models (LLMs) by guiding their step-by-step reasoning toward a final answer. However, existing PRMs either treat each reasoning step in isolation, failing to capture inter-step dependencies, or struggle to align process rewards with the final outcome. Consequently, the reward signal fails to respect temporal causality in sequential reasoning and faces ambiguous credit assign…

@arXiv_csIR_bot@mastoxiv.page
2025-10-01 09:58:57

MR$^2$-Bench: Going Beyond Matching to Reasoning in Multimodal Retrieval
Junjie Zhou, Ze Liu, Lei Xiong, Jin-Ge Yao, Yueze Wang, Shitao Xiao, Fenfen Lin, Miguel Hu Chen, Zhicheng Dou, Siqi Bao, Defu Lian, Yongping Xiong, Zheng Liu
https://arxiv.org/abs/2509.26378

MR$^2$-Bench: Going Beyond Matching to Reasoning in Multimodal Retrieval
Multimodal retrieval is becoming a crucial component of modern AI applications, yet its evaluation lags behind the demands of more realistic and challenging scenarios. Existing benchmarks primarily probe surface-level semantic correspondence (e.g., object-text matching) while failing to assess the deeper reasoning required to capture complex relationships between visual and textual information. To address this gap, we introduce MR$^2$-Bench, a reasoning-intensive benchmark for multimodal retrie…

@arXiv_astrophHE_bot@mastoxiv.page
2025-10-01 09:19:07

A fast powerful X-ray transient from tidal disruption of a white dwarf
D. -Y. Li, J. Yang, W. -D. Zhang, the Einstein Probe collaborations
https://arxiv.org/abs/2509.25877 https…

A fast powerful X-ray transient from tidal disruption of a white dwarf
Stars captured by black holes (BHs) can be tore apart by strong tidal force, producing electromagnetic flares. Some 100 tidal disruption events (TDEs) have been observed, involving invariably normal gaseous stars whose debris falls onto the BH fueling the flares over years. White dwarfs (WDs)-the most prevalent compact stars a million times denser and thus tougher than gaseous stars-can only be disrupted by intermediate-mass black holes (IMBHs) of 10^2-10^5 solar masses. WD-TDEs are predicted t…

@arXiv_csCV_bot@mastoxiv.page
2025-10-01 11:50:37

Stable Cinemetrics : Structured Taxonomy and Evaluation for Professional Video Generation
Agneet Chatterjee, Rahim Entezari, Maksym Zhuravinskyi, Maksim Lapin, Reshinth Adithyan, Amit Raj, Chitta Baral, Yezhou Yang, Varun Jampani
https://arxiv.org/abs/2509.26555