Tootfinder

Opt-in global Mastodon full text search. Join the index!

No exact results. Similar results found.
@arXiv_csCV_bot@mastoxiv.page
2025-06-09 10:08:12

STARFlow: Scaling Latent Normalizing Flows for High-resolution Image Synthesis
Jiatao Gu, Tianrong Chen, David Berthelot, Huangjie Zheng, Yuyang Wang, Ruixiang Zhang, Laurent Dinh, Miguel Angel Bautista, Josh Susskind, Shuangfei Zhai
arxiv.org/abs/2506.06276

STARFlow: Scaling Latent Normalizing Flows for High-resolution Image Synthesis
We present STARFlow, a scalable generative model based on normalizing flows that achieves strong performance in high-resolution image synthesis. The core of STARFlow is Transformer Autoregressive Flow (TARFlow), which combines the expressive power of normalizing flows with the structured modeling capabilities of Autoregressive Transformers. We first establish the theoretical universality of TARFlow for modeling continuous distributions. Building on this foundation, we introduce several key arch…

@pgcd@mastodon.online
2025-06-07 14:37:31

#happypride

Nick Offerman slams 'Parks and Rec' fan insinuating Ron Swanson is homophobic
Nick Offerman is standing up for queer 'Parks and Rec' fans after a user shared a manipulated image of his character, Ron Swanson, to spread homophobic hate: 'You dumb f---.

@gray17@mastodon.social
2025-06-07 01:42:53

teaching a transformer net to approximate ray-tracing
input is a sequence of triangle data (limited to 4k triangles), and a sequence of 8x8 radiance pixels representing the camera (limited to 512x512 resolution)
output is a sequence of 8x8 pixels of the rendered scene. training was ~10 days. generation is ~100ms (NVIDIA A100)
the results are pretty good. scaling will be a challenge (since transformers are quadratic in sequence length)

RenderFormer: Transformer-based Neural Rendering of Triangle Meshes with Global Illumination
We present RenderFormer, a neural rendering pipeline that directly renders an image from a triangle-based representation of scene with full global illumination effects, and that does not require per-scene training or finetuning.

@sillon_fictionnel@paperbay.org
2025-06-01 12:39:18

"Dans la voix, la langue résiste Š la littérature. Par cette manière de retourner la table du sens, la voix déploie tout ce qu’elle ne peut pas et l’offre négativement Š qui sait l’entendre. Avec l’incertitude de l’onde sonore qui se fracasse."
🔗 abrupt.cc/charles-pennequin/ra

sont donc lamentables, et puis nourries par des images mentales, comme quand je lisais Beckett, ou Céline: j'imagi nais des sillons profonds et noirs qui s’éloignaient ou rejoignaient le centre. Le centre était le trou. L'écriture cest comme des sillons sombres qui n’en finissent pas de se multiplier, qui n’en finissent pas de creuser et qui bordent un trou noir. Quand je lisais Antonin Artaud, c’était pareil : une image mentale de l'écriture. Mentalement, je voyais les mots qui se hérissaient …

Raconteries
Dans la voix, la langue résiste à la littérature. Par cette manière de retourner la table du sens, la voix déploie tout ce qu'elle ne peut pas et l'offre négativement à qui sait l'entendre. Avec l'incertitude de l'onde sonore qui se fracasse. Des interstices qui se veulent continents. Tendre l'oreille et observer le signe brut, creux. Lui cracher à la figure. Cracher à la figure d'un signe, c'est comme cracher à la figure d'un fantôme. C'est un geste qui nous traverse plus qu'il ne t…

@arXiv_csGR_bot@mastoxiv.page
2025-06-06 07:18:07

SSIMBaD: Sigma Scaling with SSIM-Guided Balanced Diffusion for AnimeFace Colorization
Junpyo Seo (Department of Computer Science, Seoul National University), Hanbin Koo (Department of Computer Science, Seoul National University), Jieun Yook (Department of Computer Science, Seoul National University), Byung-Ro Moon (Department of Computer Science, Seoul National University)
…

SSIMBaD: Sigma Scaling with SSIM-Guided Balanced Diffusion for AnimeFace Colorization
We propose a novel diffusion-based framework for automatic colorization of Anime-style facial sketches. Our method preserves the structural fidelity of the input sketch while effectively transferring stylistic attributes from a reference image. Unlike traditional approaches that rely on predefined noise schedules - which often compromise perceptual consistency -- our framework builds on continuous-time diffusion models and introduces SSIMBaD (Sigma Scaling with SSIM-Guided Balanced Diffusion). …

@ukraine_live_tagesschau@mastodon.social
2025-06-02 11:01:14

Selenskyj definiert Schwerpunkte der Verhandlungen
Der ukrainische Präsident Wolodymyr Selenskyj hat die Verhandlungsbereitschaft seines Landes unterstrichen. "Wir sind bereit, die nötigen Schritte für den Frieden zu unternehmen", sagte er während eines Besuchs in Litauen. Als "Ausgangspunkt" für eine Einigung nannte Selenskyj eine Waffenruhe sowie "humanitäre Maßnahmen", die Befreiung von Kriegsgef…
📷

@k8s_cncf_nbg@social.kontrollapparat.de
2025-06-03 14:11:48

Our meetup was well-attended with two interesting talks:
📋 Sarah Julia Kriesch presented "Docker vs Podman" - a beginner-friendly introduction to container basics and key differences between these tools.
🔧 Philip Laine showcased "Spegel", an OCI Registry Mirror that leverages torrent technology for efficient container image distribution - saving traffic costs and safeguarding against registry outages.
1/2

The scene depicts a bright room with cozy, modern furnishings. A person stands on the right side of the frame behind a simple podium, delivering a presentation. A slide with the title "Podman" is projected on the wall. It states that Podman is a rootless alternative to Docker, developed by Red Hat. Additional points cover the architecture, security through SELinux and AppArmor, as well as supplementary tools like Buildah, Skopeo, runc, and crun. On the left side of the room, there is a gray so…
In the same room, another person is now standing at the podium, giving a presentation. On the wall, a slide titled "$ whoami" is displayed. It features a circular photo of a person on a bicycle, along with text containing personal information: Philip Laine is of Swedish-Finnish descent, lives in Berlin, works as a developer, is the founder of the "ReFlektion" project, and was previously a maintainer of the "FLUX" tool. It’s also humorously noted that he was once nearly recruited by Microsoft. …
@arXiv_physicsinsdet_bot@mastoxiv.page
2025-06-05 07:35:00

Effect of chromatic dispersion on image size and lattice plane curvature measurements with Rocking Curve Imaging
Elena Ferrari, Claudio Ferrari, Jos\`e Baruchel, Thu Ni Tran Caliste, Stefano Basso, Akanshu Chauhan, Bianca Salmaso, Davide Sisana, Daniele Spiga
arxiv.org/abs/2506.03787…

Effect of chromatic dispersion on image size and lattice plane curvature measurements with Rocking Curve Imaging
ADP crystals of large dimensions (80x80x20 mm3), to be used as a X-ray beam expanders in the BEaTriX facility at INAF-OABrera, have been characterised at BM05 beamline at ESRF synchrotron with the main purpose to determine lattice plane curvature with a unprecedent accuracy, as the BEaTriX setup requires a radius of curvature larger than 22 km. In this beamline, the monochromator is made by 2 Si(111) parallel crystals in the non-dispersive configuration. Due to the difference in the Bragg angle…

@arXiv_physicsappph_bot@mastoxiv.page
2025-06-04 13:51:48

This arxiv.org/abs/2501.01643 has been replaced.
initial toot: mastoxiv.page/@arX…

A Polarimetry-based Field-deployable Non-interruptive Mirror Soiling Detection Method
The soiling level of heliostat mirrors in Concentrated Solar Power (CSP) fields is one of the key factors that significantly influences optical efficiency. State-of-the-art methods of monitoring heliostats soiling levels still face various challenges, including slow speed, labor-intensive operations, resolution and accuracy constraints or interruptions to solar field operations. We present a rapid, cost-effective, and non-intrusive method for mirror soiling detection based on polarimetric imagi…

@arXiv_condmatsoft_bot@mastoxiv.page
2025-06-03 07:40:25

How important is the dielectric constant in water modeling? Evaluation of the performance of the TIP4P/$\varepsilon$ force field and its compatibility with the Joung-Cheetham NaCl model
{\L}ukasz Baran, Cosmin A. Dicu-Gohoreanu, Luis G. MacDowell
arxiv.org/abs/2506.00153

How important is the dielectric constant in water modeling? Evaluation of the performance of the TIP4P/$\varepsilon$ force field and its compatibility with the Joung-Cheetham NaCl model
Efficient large scale computer simulations of complex aqueous solutions require the use of accurate but simple empirical force fields for water. However, the growing complexity of these systems is evidencing the difficulties to describe solution properties without due account of polarization. Conflicting strategies to remedy this problem are parametrizing water force fields to the dielectric constant or charge scaling of the solvated ions. In this work we compare results from TIP4P/$\varepsil…