🔥Nvidia SpatialClaw is out🔥

👉From Nvidia a novel training-free framework for spatial reasoning that adopts code as the action interface. SpatialClaw lets a VLM-backed agent write Python in a persistent kernel, composing perception modules, inspecting intermediate results, and revising its strategy across steps. Impressive: +11.2 points on 20 benchmarks💙

👉Review https://t.ly/7JB0x
👉Paper https://arxiv.org/pdf/2606.13673
👉Project https://spatialclaw.github.io/
👉Repo https://github.com/NVlabs/SpatialClaw

🏯Worldwide Semantic Facade🏯

👉A centimeter-accurate / cross-continental facade point clouds, with fine-grained semantic segmentation of architectural elements, and hierarchical facade taxonomy. 2.7B Dataset💙

👉Review https://t.ly/PpyFD
👉Paper https://arxiv.org/pdf/2607.02018
👉Project jiangyuanwangyi.github.io/UnderOneFacade_official
👉Data drive.google.com/drive/folders/1Yzz7PmyeK1qeOtkTFCfkbw7IEHXcMJo8

🐈‍⬛Spatial-perception native ViT🐈‍⬛

👉LingBot-Vision, a vision foundation model pretrained to be spatial-perception native. Better than 7x bigger foundational models. Repo under Apache💙

👉Review https://t.ly/9xIso
👉Paper https://arxiv.org/pdf/2607.05247
👉Project https://technology.robbyant.com/lingbot-vision
👉Repo https://github.com/robbyant/lingbot-vision

🏵️SoccerNet 2026 Results🏵️

👉The SoccerNet 2026 Challenges constitute the sixth annual edition of the SoccerNet open benchmarking effort, dedicated to advancing computer vision research in sports video understanding💙

👉Review https://t.ly/sfD4T
👉Paper https://lnkd.in/dSBgW_3s
👉Project https://lnkd.in/dfdmuvG8

🔥ZipDepth: Depth on Any Device🔥

👉ZipDepth from UniBO is a super-compact monocular depth network by combining an efficient reparameterizable encoder-decoder with large-scale knowledge distillation from a foundation model. Repo under MIT💙

👉Review https://t.ly/qYrLZ
👉Paper https://arxiv.org/pdf/2607.08771
👉Project https://zipdepth.github.io/
👉Repo https://github.com/fabiotosi92/ZipDepth

💋SAM-MT: Real-Time Multi-Target VOS💋

👉Fudan & Shangai unveil SAM-MT, an efficient interactive multi-target video segmentation framework that maintains near-single-object efficiency (FPS/VRAM) as target count increases, while maintaining robust video segmentation performance. Repo available💙

👉Review https://t.ly/Z_4C7
👉Paper https://lnkd.in/dvS-iyBD
👉Project https://lnkd.in/daQ8na8T
👉Repo https://lnkd.in/dgbX2tZv

🌔Foundation Global SFM🌔

👉Glob3R is a global SfM-style reconstruction built on 3D foundation models. key idea: explicitly optimize feed-forward geometric predictions. Repo TBA💙

👉Review https://t.ly/Z_4C7
👉Paper https://arxiv.org/pdf/2607.09225
👉Project https://junyuandeng.github.io/Glob3r/
👉Repo TBA

🎂REMIND: long-term MOT re-ID🎂

👉REMIND by CVAR-UPM is a novel online tracker designed for long-term multi-object re-ID of generic indoor objects from monocular RGB, requiring neither camera pose nor depth. Repo under MIT💙

👉Review https://t.ly/AkQoI
👉Paper https://lnkd.in/dm58mkCv
👉Project https://lnkd.in/dZrAZqFe
👉Repo https://lnkd.in/dbidrwxU

🦧 MonkeyOCRv2 is out! 🦧

👉MonkeyOCRv2 is a text-centric visual foundation model that unifies fine-grained text modeling, cross-task representation learning, and cross-lingual generalization in a single encoder. Released for academic research and non-commercial use💙

👉Review https://t.ly/yicEK
👉Paper https://arxiv.org/pdf/2607.11562
👉Repo https://github.com/Yuliang-Liu/MonkeyOCRv2

🌈FlowWAM: flow->action prediction🌈

👉FlowWAM is a novel dual-stream diffusion framework that adopts optical flow as a unified, video-native action representation. Repo under Apache💙

👉Review https://t.ly/FmutT
👉Paper https://arxiv.org/abs/2607.13017
👉Project https://flow-wam.github.io/
👉Repo github.com/YixiangChen515/FlowWAM

🏯SOTA Music-to-Dance Gen🏯

👉The Tongyi Lab unveils Wan-Dancer, a novel stable minute-scale synthesis at 720p/30fps across five dance genres. Impressive results, new SOTA on long clip by a large margin. Repo under Apache 2.0💙

👉Review https://t.ly/AKY5j
👉Paper https://lnkd.in/d_xA7dwb
👉Project https://lnkd.in/dzfnw2h4
👉Repo https://lnkd.in/d-Zj_cTf

👉Not a render. Not a concept. This is GENE.01 by Generative Bionics, the Italians coolest scaleup strikes back: in just six months, they turned GENE.01 into a fully functional humanoid platform that can walk, sense and interact.

👉Full-body multimodal skin perceives touch, proximity, force and temperature, bringing Physical AI closer to safe and natural collaboration with people.

👉More: https://t.ly/F3I3A