Differentiable NMS via Sinkhorn Matching for End-to-End Fabric Defect Detection

Posted by dbos Berkeley

Kavli Affiliate: Feng Wang | First 5 Authors: Zhengyang Lu, Bingjie Lu, Weifan Wang, Feng Wang, | Summary: Fabric defect detection confronts two fundamental challenges. First, conventional non-maximum suppression disrupts gradient flow, which hinders genuine end-to-end learning. Second, acquiring pixel-level annotations at industrial scale is prohibitively costly. Addressing these limitations, we propose a differentiable NMS […]


Continue.. Differentiable NMS via Sinkhorn Matching for End-to-End Fabric Defect Detection

Compilation strategies for quantum network programs using Qoala

Posted by dbos Delft

Kavli Affiliate: Stephanie Wehner | First 5 Authors: Samuel Oslovich, Bart van der Vecht, Stephanie Wehner, , | Summary: Execution of quantum network applications requires a software stack for nodes. Recently, the first designs and demonstrations have been proposed for such software stacks, including QNodeOS and its extension, Qoala. The latter enables compilation strategies previously […]


Continue.. Compilation strategies for quantum network programs using Qoala

Inferring charge noise source locations from correlations in spin qubits

Posted by dbos Delft

Kavli Affiliate: Giordano Scappucci | First 5 Authors: Juan S. Rojas-Arias, Akito Noiri, Jun Yoneda, Peter Stano, Takashi Nakajima | Summary: We investigate low-frequency noise in a spin-qubit device made in isotopically purified Si/Si-Ge. Observing sizable cross-correlations among energy fluctuations of different qubits, we conclude that these fluctuations are dominated by charge noise. At low […]


Continue.. Inferring charge noise source locations from correlations in spin qubits

SmartTrap: Automated Precision Experiments with Optical Tweezers

Posted by dbos Berkeley

Kavli Affiliate: Carlos Bustamante | First 5 Authors: Martin Selin, Antonio Ciarlo, Giuseppe Pesce, Lars Bengtsson, Joan Camunas-Soler | Summary: There is a trend in research towards more automation using smart systems powered by artificial intelligence. While experiments are often challenging to automate, they can greatly benefit from automation by reducing labor and increasing reproducibility. […]


Continue.. SmartTrap: Automated Precision Experiments with Optical Tweezers

Lay-Your-Scene: Natural Scene Layout Generation with Diffusion Transformers

Posted by dbos Berkeley

Kavli Affiliate: Xiang Zhang | First 5 Authors: Divyansh Srivastava, Xiang Zhang, He Wen, Chenru Wen, Zhuowen Tu | Summary: We present Lay-Your-Scene (shorthand LayouSyn), a novel text-to-layout generation pipeline for natural scenes. Prior scene layout generation methods are either closed-vocabulary or use proprietary large language models for open-vocabulary generation, limiting their modeling capabilities and […]


Continue.. Lay-Your-Scene: Natural Scene Layout Generation with Diffusion Transformers

Nanoscale Mechanical Structures Fabricated from Silicon-on-Insulator Substrates

Posted by dbos Caltech

Kavli Affiliate: Michael L. Roukes | First 5 Authors: Andrew N. Cleland, Michael L. Roukes, , , | Summary: We describe a method with which to fabricate sub-micron mechanical structures from silicon-on-insulator substrates. We believe this is the first reported method for such fabrication, and our technique allows for complex, multilayer electron beam lithography to […]


Continue.. Nanoscale Mechanical Structures Fabricated from Silicon-on-Insulator Substrates

Spalled barium titanate single crystal thin films for functional device applications

Posted by dbos Caltech

Kavli Affiliate: Harry A. Atwater | First 5 Authors: Prachi Thureja, Andrew W. Nyholm, Martin Thomaschewski, Phillip R. Jahelka, Julie Belleville | Summary: We report a scalable approach for fabricating single-crystal barium titanate (BTO) thin films through spalling from bulk substrates. Conventional thin film growth techniques often face challenges in achieving high-quality single crystal microstructure […]


Continue.. Spalled barium titanate single crystal thin films for functional device applications

MeshGen: Generating PBR Textured Mesh with Render-Enhanced Auto-Encoder and Generative Data Augmentation

Posted by dbos Berkeley

Kavli Affiliate: Feng Wang | First 5 Authors: Zilong Chen, Yikai Wang, Wenqiang Sun, Feng Wang, Yiwen Chen | Summary: In this paper, we introduce MeshGen, an advanced image-to-3D pipeline that generates high-quality 3D meshes with detailed geometry and physically based rendering (PBR) textures. Addressing the challenges faced by existing 3D native diffusion models, such […]


Continue.. MeshGen: Generating PBR Textured Mesh with Render-Enhanced Auto-Encoder and Generative Data Augmentation

The True Stellar Obliquity of a Sub-Saturn Planet from the Tierras Observatory and KPF

Posted by dbos Harvard

Kavli Affiliate: David Charbonneau | First 5 Authors: Patrick Tamburo, Samuel W. Yee, Juliana García-Mejía, Gudmundur Stefánsson, David Charbonneau | Summary: We measure the true obliquity of TOI-2364, a K dwarf with a sub-Saturn-mass ($M_p = 0.18,M_J$) transiting planet on the upper edge of the hot Neptune desert. We used new Rossiter-McLaughlin observations gathered with […]


Continue.. The True Stellar Obliquity of a Sub-Saturn Planet from the Tierras Observatory and KPF

DexCtrl: Towards Sim-to-Real Dexterity with Adaptive Controller Learning

Posted by dbos Berkeley

Kavli Affiliate: Xiang Zhang | First 5 Authors: Shuqi Zhao, Ke Yang, Yuxin Chen, Chenran Li, Yichen Xie | Summary: Dexterous manipulation has seen remarkable progress in recent years, with policies capable of executing many complex and contact-rich tasks in simulation. However, transferring these policies from simulation to real world remains a significant challenge. One […]


Continue.. DexCtrl: Towards Sim-to-Real Dexterity with Adaptive Controller Learning