Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

We’re introducing Segmentation. SVG masks from prompt, points, or box. SOTA on benchmarks.

moondreamai's profile picture

moondream

12,221 subscribers

153,393 görüntüleme • 8 ay önce •via X (Twitter)

EğitimHaberler & PolitikaBilim & Teknoloji
LIVE

Anya Rossi• Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

Segment Any 3D Gaussians paper page: Interactive 3D segmentation in radiance fields is an appealing task since its importance in 3D scene understanding and manipulation. However, existing methods face challenges in either achieving fine-grained, multi-granularity segmentation or contending with substantial computational overhead, inhibiting real-time interaction. In this paper, we introduce Segment Any 3D GAussians (SAGA), a novel 3D interactive segmentation approach that seamlessly blends a 2D segmentation foundation model with 3D Gaussian Splatting (3DGS), a recent breakthrough of radiance fields. SAGA efficiently embeds multi-granularity 2D segmentation results generated by the segmentation foundation model into 3D Gaussian point features through well-designed contrastive training. Evaluation on existing benchmarks demonstrates that SAGA can achieve competitive performance with state-of-the-art methods. Moreover, SAGA achieves multi-granularity segmentation and accommodates various prompts, including points, scribbles, and 2D masks. Notably, SAGA can finish the 3D segmentation within milliseconds, achieving nearly 1000x acceleration compared to previous SOTA.
0:18

Segment Any 3D Gaussians paper page: Interactive 3D segmentation in radiance fields is an appealing task since its importance in 3D scene understanding and manipulation. However, existing methods face challenges in either achieving fine-grained, multi-granularity segmentation or contending with substantial computational overhead, inhibiting real-time interaction. In this paper, we introduce Segment Any 3D GAussians (SAGA), a novel 3D interactive segmentation approach that seamlessly blends a 2D segmentation foundation model with 3D Gaussian Splatting (3DGS), a recent breakthrough of radiance fields. SAGA efficiently embeds multi-granularity 2D segmentation results generated by the segmentation foundation model into 3D Gaussian point features through well-designed contrastive training. Evaluation on existing benchmarks demonstrates that SAGA can achieve competitive performance with state-of-the-art methods. Moreover, SAGA achieves multi-granularity segmentation and accommodates various prompts, including points, scribbles, and 2D masks. Notably, SAGA can finish the 3D segmentation within milliseconds, achieving nearly 1000x acceleration compared to previous SOTA.

AK

69,542 görüntüleme • 2 yıl önce

Moondream segmenting delivers pixel-accurate vector masks for arms, packages, and conveyor bins. A free, open-source model with state-of-the-art segmentation benchmarks, built for real-world automation.
0:11

Moondream segmenting delivers pixel-accurate vector masks for arms, packages, and conveyor bins. A free, open-source model with state-of-the-art segmentation benchmarks, built for real-world automation.

moondream

15,141 görüntüleme • 7 ay önce

Introducing benchmark tracking across all of arXiv With thousands of models and benchmarks out there, it's impossible to track what's actually state-of-the-art We now index results from millions of papers to surface which models are SOTA according to open source research!
0:17

Introducing benchmark tracking across all of arXiv With thousands of models and benchmarks out there, it's impossible to track what's actually state-of-the-art We now index results from millions of papers to surface which models are SOTA according to open source research!

alphaXiv

10,892 görüntüleme • 7 ay önce

Grep just achieved SOTA on the three major deep research benchmarks, beating Perplexity, Google, Nvidia, OpenAI, and Anthropic. We're a two-person founding team.
0:15

Grep just achieved SOTA on the three major deep research benchmarks, beating Perplexity, Google, Nvidia, OpenAI, and Anthropic. We're a two-person founding team.

AJ Asver

64,500 görüntüleme • 3 ay önce

Today we're releasing the Segment Anything Model (SAM) — a step toward the first foundation model for image segmentation. SAM is capable of one-click segmentation of any object from any photo or video + zero-shot transfer to other segmentation tasks ➡️
0:15

Today we're releasing the Segment Anything Model (SAM) — a step toward the first foundation model for image segmentation. SAM is capable of one-click segmentation of any object from any photo or video + zero-shot transfer to other segmentation tasks ➡️

AI at Meta

3,570,780 görüntüleme • 3 yıl önce

Most 3DGS segmentation tools either pre‑train per scene or lock errors into a feature field you can’t undo. ArtisanGS instead turns a few 2D masks into editable 3D object selections via Cutie tracking + black‑box splat aggregation, then lets you iteratively correct mistakes with consistent 2D/3D selection modes. 📗 #NVIDIAResearch paper:
0:15

Most 3DGS segmentation tools either pre‑train per scene or lock errors into a feature field you can’t undo. ArtisanGS instead turns a few 2D masks into editable 3D object selections via Cutie tracking + black‑box splat aggregation, then lets you iteratively correct mistakes with consistent 2D/3D selection modes. 📗 #NVIDIAResearch paper:

NVIDIA AI Developer

42,369 görüntüleme • 5 ay önce

This might be big Instead of just sending one prompt to multiple agents and comparing the results manually... this D1 layer is: - Deciding which agent acts at each step - Only 3B active parameters - SOTA on router benchmarks So it just acts between you and the different agents Pretty cool!
0:11

This might be big Instead of just sending one prompt to multiple agents and comparing the results manually... this D1 layer is: - Deciding which agent acts at each step - Only 3B active parameters - SOTA on router benchmarks So it just acts between you and the different agents Pretty cool!

Paul Couvert

10,523 görüntüleme • 12 gün önce

you can now generate prompts from images in Krea drag & drop any image into the prompt box -> get a prompt for it
0:18

you can now generate prompts from images in Krea drag & drop any image into the prompt box -> get a prompt for it

titus

16,886 görüntüleme • 5 ay önce

just updated my Segment Anything 2 (SAM2) Hugging Face space now you can run the model in two modes - box prompt and automatic mask generation; I'll try to add short video segmentation today link to space if you want to have some fun:
0:15

just updated my Segment Anything 2 (SAM2) Hugging Face space now you can run the model in two modes - box prompt and automatic mask generation; I'll try to add short video segmentation today link to space if you want to have some fun:

SkalskiP

16,075 görüntüleme • 2 yıl önce

The new model Extract Object by Bria is now live on fal! - Isolates any object from an image with a single API call - Turns a natural-language prompt into a clean RGBA-PNG cutout - Outputs composite-ready results with no masks or manual editing required
0:10

The new model Extract Object by Bria is now live on fal! - Isolates any object from an image with a single API call - Turns a natural-language prompt into a clean RGBA-PNG cutout - Outputs composite-ready results with no masks or manual editing required

fal

33,631 görüntüleme • 1 ay önce

Screen Studio 3.1 is here Introducing Masks: Blur sensitive data or highlight what matters in your recordings. From passwords to API keys - keep your data private while recording. Give it a try and share your feedback!
0:18

Screen Studio 3.1 is here Introducing Masks: Blur sensitive data or highlight what matters in your recordings. From passwords to API keys - keep your data private while recording. Give it a try and share your feedback!

Screen Studio

108,702 görüntüleme • 1 yıl önce

.Groq Inc has introduced prompt caching for the SOTA open-source coding model, Kimi-k2. What does this mean for you? Significantly reduced costs with Kimi-k2 on its fastest provider. Check the price: left side without prompt caching, right side with prompt caching.
0:15

.Groq Inc has introduced prompt caching for the SOTA open-source coding model, Kimi-k2. What does this mean for you? Significantly reduced costs with Kimi-k2 on its fastest provider. Check the price: left side without prompt caching, right side with prompt caching.

Cline

29,843 görüntüleme • 11 ay önce

RF-DETR just landed to Hugging Face transformers 🥵🔥 sota real-time detection & segmentation models by Roboflow 💜 > play with our real-time demo > fine-tune the models on your use case with our tutorials (takes a toaster's VRAM) > or just hand them to your agents 😄
0:13

RF-DETR just landed to Hugging Face transformers 🥵🔥 sota real-time detection & segmentation models by Roboflow 💜 > play with our real-time demo > fine-tune the models on your use case with our tutorials (takes a toaster's VRAM) > or just hand them to your agents 😄

merve

56,967 görüntüleme • 2 ay önce

🚀A Simple Baseline for Streaming Video Understanding🚀 #SimpleStream reveals a simple baseline of *recent N-frame sliding window* beats SOTA memory-based methods on standard streaming video benchmarks lmms-lab - Project: - Code:
0:19

🚀A Simple Baseline for Streaming Video Understanding🚀 #SimpleStream reveals a simple baseline of *recent N-frame sliding window* beats SOTA memory-based methods on standard streaming video benchmarks lmms-lab - Project: - Code:

Ziwei Liu

36,089 görüntüleme • 4 ay önce

Introducing the Moneyball Card! 🏀 Each 3-pointer = +4 Points Splash from deep, stack the bag 💰
0:19

Introducing the Moneyball Card! 🏀 Each 3-pointer = +4 Points Splash from deep, stack the bag 💰

Chalkboard

188,216 görüntüleme • 3 ay önce

introducing Scan 🩻 give a prompt to claude or codex and it makes this beautiful shareable maps of your codebase free, fast, and oss!
0:18

introducing Scan 🩻 give a prompt to claude or codex and it makes this beautiful shareable maps of your codebase free, fast, and oss!

Gustavo

342,848 görüntüleme • 21 gün önce

Bridgeless BTC staking with triple points from @SymbioticFi, @MellowProtocol, & iBTC—it's here 🔥 We’re live on the Symbiotic mainnet with a new Mellow vault at launch More yield, more points👇
0:18

Bridgeless BTC staking with triple points from @SymbioticFi, @MellowProtocol, & iBTC—it's here 🔥 We’re live on the Symbiotic mainnet with a new Mellow vault at launch More yield, more points👇

BitSafe

10,382 görüntüleme • 1 yıl önce

Introducing Sleap Points. Unlock Rewards from your Loyalty.🤝 Earn Points from your hotel bookings.🏨 Multiply your points by holding city NFTs.🌆 Elevate your travel experience one trip at a time. More details coming soon. Q1 2025.🛎️
0:14

Introducing Sleap Points. Unlock Rewards from your Loyalty.🤝 Earn Points from your hotel bookings.🏨 Multiply your points by holding city NFTs.🌆 Elevate your travel experience one trip at a time. More details coming soon. Q1 2025.🛎️

Sleap

19,939 görüntüleme • 1 yıl önce

we built a new way to index and chat with any PDF. introducing tree-guided hybrid agentic search. Nozomio Labs indexes PDFs into document trees, reasons over structure, and uses agents to retrieve precise answers in minutes. we’re seeing strong performance on financebench and other benchmarks. curious what you think!
0:19

we built a new way to index and chat with any PDF. introducing tree-guided hybrid agentic search. Nozomio Labs indexes PDFs into document trees, reasons over structure, and uses agents to retrieve precise answers in minutes. we’re seeing strong performance on financebench and other benchmarks. curious what you think!

Arlan

26,209 görüntüleme • 6 ay önce

We’re introducing new steps to improve grassroots youth football from the 2026/27 season. Find out more ⬇️
0:15

We’re introducing new steps to improve grassroots youth football from the 2026/27 season. Find out more ⬇️

England Football

184,224 görüntüleme • 1 yıl önce