Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Excited to share Spotiflow, our new spot detection method for image-based spatial transcriptomics that facilitates the analysis of large iST images. Lead by Albert Dominguez Mantes jointly w Gioele La Manno EPFL Life Sciences EPFL Center for Imaging EPFL Institute of Bioengineering

martweig's profile picture

Martin Weigert (maweigert.bsky.social)

1,739 subscribers

25,696 Aufrufe • vor 2 Jahren •via X (Twitter)

BildungWissenschaft & Technologie
LIVE

Anya Rossi• Live Now

Private livecam show

3 Kommentare

Profilbild von Martin Weigert (maweigert.bsky.social)
Martin Weigert (maweigert.bsky.social)vor 2 Jahren

Spotiflow uses multiscale Gaussian heatmap prediction alongside a unique stereographic flow regression, which smoothly embeds the local closest-point vector field onto the unit sphere, thereby avoiding the ambiguity at infinity

Profilbild von Martin Weigert (maweigert.bsky.social)
Martin Weigert (maweigert.bsky.social)vor 2 Jahren

Spotiflow achieves SOTA results on 7 synthetic and real datasets while being faster than commonly used iST spot detection methods. It comes with pre-trained models, a python library, and a plugin for @napari_imaging

Profilbild von Martin Weigert (maweigert.bsky.social)
Martin Weigert (maweigert.bsky.social)vor 2 Jahren

Thanks to the many amazing people at @epflSV that contributed: Toni Herrera, Irina Khven, @CanAztekin, @Anjalie_Sch, Georgios Tsissios, Eftychia Kyriacou, Joachim Liger. Feedback welcome!

Ähnliche Videos

A drone with avian-inspired legs by EPFL.
0:40

A drone with avian-inspired legs by EPFL.

MachinePix

67,054 Aufrufe • vor 1 Jahr

It's annoying when your shoelaces become undone, but no big problem; just retie them. But what about surgical knots? What sets their strength? Why do surgeons take so long to learn how to tie safe knots? Check out our recent study EPFL School of Engineering EPFL
0:37

It's annoying when your shoelaces become undone, but no big problem; just retie them. But what about surgical knots? What sets their strength? Why do surgeons take so long to learn how to tie safe knots? Check out our recent study EPFL School of Engineering EPFL

Flexible Structures Laboratory @ EPFL

14,651 Aufrufe • vor 3 Jahren

EPFL researchers have built a drone that can walk, hop, and jump into flight with the aid of birdlike legs, greatly expanding the range of potential environments accessible to unmanned aerial vehicles. RAVEN (Robotic Avian-inspired Vehicle for multiple ENvironments), designed based on perching birds like ravens and crows that frequently switch between air and land, the multifunctional robotic legs allow it to take off autonomously in environments previously inaccessible to winged drones. Check out the full article here: Image/video credits: EPFL / Ulrich M. / Akshet Patel #drones #robotics #technology #engineering #stem #innovation
1:00

EPFL researchers have built a drone that can walk, hop, and jump into flight with the aid of birdlike legs, greatly expanding the range of potential environments accessible to unmanned aerial vehicles. RAVEN (Robotic Avian-inspired Vehicle for multiple ENvironments), designed based on perching birds like ravens and crows that frequently switch between air and land, the multifunctional robotic legs allow it to take off autonomously in environments previously inaccessible to winged drones. Check out the full article here: Image/video credits: EPFL / Ulrich M. / Akshet Patel #drones #robotics #technology #engineering #stem #innovation

Wevolver

37,365 Aufrufe • vor 1 Jahr

EPFL researchers have built a drone that can walk, hop, and jump into flight with the aid of birdlike legs, greatly expanding the range of potential environments accessible to unmanned aerial vehicles. RAVEN (Robotic Avian-inspired Vehicle for multiple ENvironments), designed based on perching birds like ravens and crows that frequently switch between air and land, the multifunctional robotic legs allow it to take off autonomously in environments previously inaccessible to winged drones. Image/video credits: EPFL / Ulrich M. / Akshet Patel #drones #robotics #technology #engineering #stem #innovation -------------------------------- Stay ahead of the curve! Follow us now on our WhatsApp ( and Telegram ( channels and stay updated about the cutting edge.
1:00

EPFL researchers have built a drone that can walk, hop, and jump into flight with the aid of birdlike legs, greatly expanding the range of potential environments accessible to unmanned aerial vehicles. RAVEN (Robotic Avian-inspired Vehicle for multiple ENvironments), designed based on perching birds like ravens and crows that frequently switch between air and land, the multifunctional robotic legs allow it to take off autonomously in environments previously inaccessible to winged drones. Image/video credits: EPFL / Ulrich M. / Akshet Patel #drones #robotics #technology #engineering #stem #innovation -------------------------------- Stay ahead of the curve! Follow us now on our WhatsApp ( and Telegram ( channels and stay updated about the cutting edge.

Wevolver

22,013 Aufrufe • vor 11 Monaten

New @epfl paper on inflatables to be presented by @samara_ren at ACM SIGGRAPH in August. More info at: Joint work with Julian Panetta, Seiichi Suzuki, Uday Kusupati, Florin Isvoranu #inflatables #design #architecture #optimization
0:36

New @epfl paper on inflatables to be presented by @samara_ren at ACM SIGGRAPH in August. More info at: Joint work with Julian Panetta, Seiichi Suzuki, Uday Kusupati, Florin Isvoranu #inflatables #design #architecture #optimization

GCM

18,723 Aufrufe • vor 2 Jahren

Morpho is a power plant inspection drone developed by Elythor, an EPFL spinoff company. It uses wind to extend flight time and has adjustable wings for vertical and horizontal flight. Used to inspect power plants, power lines, turbines, pipelines, and offshore platforms.
0:27

Morpho is a power plant inspection drone developed by Elythor, an EPFL spinoff company. It uses wind to extend flight time and has adjustable wings for vertical and horizontal flight. Used to inspect power plants, power lines, turbines, pipelines, and offshore platforms.

Space and Technology

129,776 Aufrufe • vor 24 Tagen

Researchers have developed fiber-like pumps that allow high-pressure fluidic circuits to be woven into textiles without an external pump. Soft supportive exoskeletons, thermoregulatory clothing, and immersive haptics can therefore be powered from pumps sewn into the fabric of the devices themselves. Credit: EPFL
0:53

Researchers have developed fiber-like pumps that allow high-pressure fluidic circuits to be woven into textiles without an external pump. Soft supportive exoskeletons, thermoregulatory clothing, and immersive haptics can therefore be powered from pumps sewn into the fabric of the devices themselves. Credit: EPFL

Wevolver

15,612 Aufrufe • vor 1 Jahr

Thrilled to share our work showing protective retinal #microglia in retinal degenerative disease out now Journal of Experimental Medicine from Duke Eye Center DukeIntegrativeImmunobiology Duke Institute for Brain Sciences ~Thank you Bex for this killer animation
2:14

Thrilled to share our work showing protective retinal #microglia in retinal degenerative disease out now Journal of Experimental Medicine from Duke Eye Center DukeIntegrativeImmunobiology Duke Institute for Brain Sciences ~Thank you Bex for this killer animation

Danny Saban

16,528 Aufrufe • vor 2 Jahren

Researchers at EPFL in Switzerland have developed a bioinspired robot called GOAT that can change its shape to adapt to different environments. It moves by adjusting the length of its cables, allowing it to morph and travel efficiently across various terrains. This shape-shifting ability makes it useful for exploration and rescue missions, and it was introduced in 2025.
0:29

Researchers at EPFL in Switzerland have developed a bioinspired robot called GOAT that can change its shape to adapt to different environments. It moves by adjusting the length of its cables, allowing it to morph and travel efficiently across various terrains. This shape-shifting ability makes it useful for exploration and rescue missions, and it was introduced in 2025.

Space and Technology

14,170 Aufrufe • vor 2 Monaten

SP80 team aims for a record with a kite-powered sailboat, targeting 150 km/h. This EPFL-driven project is in its 5th year, focusing on wind-propelled speed. 🌬️⛵🏁 #CES2024
0:57

SP80 team aims for a record with a kite-powered sailboat, targeting 150 km/h. This EPFL-driven project is in its 5th year, focusing on wind-propelled speed. 🌬️⛵🏁 #CES2024

Tansu Yegen

36,766 Aufrufe • vor 2 Jahren

Five years ago, if someone had said my final task at EPFL would involve prototyping a UI for molecular dynamics in AR, I would've called them crazy. And yet, here we are, pushing the limits of what a web browser can do! Built with Three.js and UIKit by Poimandres
0:31

Five years ago, if someone had said my final task at EPFL would involve prototyping a UI for molecular dynamics in AR, I would've called them crazy. And yet, here we are, pushing the limits of what a web browser can do! Built with Three.js and UIKit by Poimandres

Fabio Cortes

16,837 Aufrufe • vor 1 Jahr

🔥😀Super-super excited to report the publication of "Single-cell multi-omic and spatial profiling of human kidneys implicates the fibrotic microenvironment in kidney disease progression" in Nature Genetics by Amin Abedini and Jon Levinsohn #kidney #diabetes #spatial transcriptomics
1:21

🔥😀Super-super excited to report the publication of "Single-cell multi-omic and spatial profiling of human kidneys implicates the fibrotic microenvironment in kidney disease progression" in Nature Genetics by Amin Abedini and Jon Levinsohn #kidney #diabetes #spatial transcriptomics

Katalin Susztak

16,259 Aufrufe • vor 1 Jahr

We are excited to share our new version of Epitools for napari! We have developed it together with UCL Advanced Research Computing Centre and thanks to Biohub funding. UCL Faculty of Life Sciences UCL IPLS @PVicenteMunuera Giulia Paci UCL Laboratory for Molecular Cell Biology #Epitools #opensource
1:14

We are excited to share our new version of Epitools for napari! We have developed it together with UCL Advanced Research Computing Centre and thanks to Biohub funding. UCL Faculty of Life Sciences UCL IPLS @PVicenteMunuera Giulia Paci UCL Laboratory for Molecular Cell Biology #Epitools #opensource

Yanlan Mao

14,147 Aufrufe • vor 2 Jahren

Excited to share our work on Neural Assets: a new method for enabling 3D asset-level control in image diffusion models – scalable & without any 3D inductive biases. Neural Assets goes beyond text or pixel-based control & provides an interface inspired by 3D graphics tools. 🧵
0:30

Excited to share our work on Neural Assets: a new method for enabling 3D asset-level control in image diffusion models – scalable & without any 3D inductive biases. Neural Assets goes beyond text or pixel-based control & provides an interface inspired by 3D graphics tools. 🧵

Thomas Kipf

97,924 Aufrufe • vor 2 Jahren

Collaborative Score Distillation for Consistent Visual Synthesis paper page: Generative priors of large-scale text-to-image diffusion models enable a wide range of new generation and editing applications on diverse visual modalities. However, when adapting these priors to complex visual modalities, often represented as multiple images (e.g., video), achieving consistency across a set of images is challenging. In this paper, we address this challenge with a novel method, Collaborative Score Distillation (CSD). CSD is based on the Stein Variational Gradient Descent (SVGD). Specifically, we propose to consider multiple samples as "particles" in the SVGD update and combine their score functions to distill generative priors over a set of images synchronously. Thus, CSD facilitates seamless integration of information across 2D images, leading to a consistent visual synthesis across multiple samples. We show the effectiveness of CSD in a variety of tasks, encompassing the visual editing of panorama images, videos, and 3D scenes. Our results underline the competency of CSD as a versatile method for enhancing inter-sample consistency, thereby broadening the applicability of text-to-image diffusion models.
1:55

Collaborative Score Distillation for Consistent Visual Synthesis paper page: Generative priors of large-scale text-to-image diffusion models enable a wide range of new generation and editing applications on diverse visual modalities. However, when adapting these priors to complex visual modalities, often represented as multiple images (e.g., video), achieving consistency across a set of images is challenging. In this paper, we address this challenge with a novel method, Collaborative Score Distillation (CSD). CSD is based on the Stein Variational Gradient Descent (SVGD). Specifically, we propose to consider multiple samples as "particles" in the SVGD update and combine their score functions to distill generative priors over a set of images synchronously. Thus, CSD facilitates seamless integration of information across 2D images, leading to a consistent visual synthesis across multiple samples. We show the effectiveness of CSD in a variety of tasks, encompassing the visual editing of panorama images, videos, and 3D scenes. Our results underline the competency of CSD as a versatile method for enhancing inter-sample consistency, thereby broadening the applicability of text-to-image diffusion models.

AK

33,500 Aufrufe • vor 2 Jahren

holy shit a team at EPFL in switzerland just published a paper where they restored walking in people with spinal cord injuries using deep brain stimulation (DBS). this isn’t just nerves reconnecting. they targeted a new brain region for movement the results are insane 1/
2:30

holy shit a team at EPFL in switzerland just published a paper where they restored walking in people with spinal cord injuries using deep brain stimulation (DBS). this isn’t just nerves reconnecting. they targeted a new brain region for movement the results are insane 1/

vittorio

902,628 Aufrufe • vor 1 Jahr

With our support & Forensic Architecture our Ukrainian partners at Center for Spatial Technologies Center for Spatial Technologies released new additions to their spatial archive, gathering digital & visual evidence & testimonies from the Mariupol Theatre bombing 2 years ago.
0:40

With our support & Forensic Architecture our Ukrainian partners at Center for Spatial Technologies Center for Spatial Technologies released new additions to their spatial archive, gathering digital & visual evidence & testimonies from the Mariupol Theatre bombing 2 years ago.

Forensis

14,694 Aufrufe • vor 2 Jahren

📢Excited to share our work on #GenerativeAI to explore disparities in large imaging datasets. 🧵(1/5) Read along for a high-level overview, -or- 🎧Listen to the #NotebookLM version. 👇
8:08

📢Excited to share our work on #GenerativeAI to explore disparities in large imaging datasets. 🧵(1/5) Read along for a high-level overview, -or- 🎧Listen to the #NotebookLM version. 👇

Bardia Khosravi, MD

10,753 Aufrufe • vor 1 Jahr

Doppelgangers: Learning to Disambiguate Images of Similar Structures paper page: We consider the visual disambiguation task of determining whether a pair of visually similar images depict the same or distinct 3D surfaces (e.g., the same or opposite sides of a symmetric building). Illusory image matches, where two images observe distinct but visually similar 3D surfaces, can be challenging for humans to differentiate, and can also lead 3D reconstruction algorithms to produce erroneous results. We propose a learning-based approach to visual disambiguation, formulating it as a binary classification task on image pairs. To that end, we introduce a new dataset for this problem, Doppelgangers, which includes image pairs of similar structures with ground truth labels. We also design a network architecture that takes the spatial distribution of local keypoints and matches as input, allowing for better reasoning about both local and global cues. Our evaluation shows that our method can distinguish illusory matches in difficult cases, and can be integrated into SfM pipelines to produce correct, disambiguated 3D reconstructions.
0:22

Doppelgangers: Learning to Disambiguate Images of Similar Structures paper page: We consider the visual disambiguation task of determining whether a pair of visually similar images depict the same or distinct 3D surfaces (e.g., the same or opposite sides of a symmetric building). Illusory image matches, where two images observe distinct but visually similar 3D surfaces, can be challenging for humans to differentiate, and can also lead 3D reconstruction algorithms to produce erroneous results. We propose a learning-based approach to visual disambiguation, formulating it as a binary classification task on image pairs. To that end, we introduce a new dataset for this problem, Doppelgangers, which includes image pairs of similar structures with ground truth labels. We also design a network architecture that takes the spatial distribution of local keypoints and matches as input, allowing for better reasoning about both local and global cues. Our evaluation shows that our method can distinguish illusory matches in difficult cases, and can be integrated into SfM pipelines to produce correct, disambiguated 3D reconstructions.

AK

76,878 Aufrufe • vor 2 Jahren

Inspired by some gesture-based point cloud controllers I've seen on here, I vibe coded a similar web app to explore the relationship between spatial, UMAP, and PCA embeddings for spatial transcriptomics data. Next level interactivity via🖐️ Try it out:
0:42

Inspired by some gesture-based point cloud controllers I've seen on here, I vibe coded a similar web app to explore the relationship between spatial, UMAP, and PCA embeddings for spatial transcriptomics data. Next level interactivity via🖐️ Try it out:

Dr. Jean Fan

17,620 Aufrufe • vor 5 Monaten