Half-Wormholes and Ensemble Averages

Kavli Affiliate: Cheng Peng | First 5 Authors: Cheng Peng, Jia Tian, Yingyu Yang, , | Summary: We study "half-wormhole-like" saddle point contributions to spectral correlators in a variety of ensemble average models, including various statistical models, generalized 0d SYK models, 1d Brownian SYK models and an extension of it. In statistical ensemble models, where […]


Continue.. Half-Wormholes and Ensemble Averages

DDDM: a Brain-Inspired Framework for Robust Classification

Kavli Affiliate: Yi Zhou | First 5 Authors: Xiyuan Chen, Xingyu Li, Yi Zhou, Tianming Yang, | Summary: Despite their outstanding performance in a broad spectrum of real-world tasks, deep artificial neural networks are sensitive to input noises, particularly adversarial perturbations. On the contrary, human and animal brains are much less vulnerable. In contrast to […]


Continue.. DDDM: a Brain-Inspired Framework for Robust Classification

SHAPE: An Unified Approach to Evaluate the Contribution and Cooperation of Individual Modalities

Kavli Affiliate: Yi Zhou | First 5 Authors: Pengbo Hu, Xingyu Li, Yi Zhou, , | Summary: As deep learning advances, there is an ever-growing demand for models capable of synthesizing information from multi-modal resources to address the complex tasks raised from real-life applications. Recently, many large multi-modal datasets have been collected, on which researchers […]


Continue.. SHAPE: An Unified Approach to Evaluate the Contribution and Cooperation of Individual Modalities

Reliable Label Correction is a Good Booster When Learning with Extremely Noisy Labels

Kavli Affiliate: Zheng Zhu | First 5 Authors: Kai Wang, Xiangyu Peng, Shuo Yang, Jianfei Yang, Zheng Zhu | Summary: Learning with noisy labels has aroused much research interest since data annotations, especially for large-scale datasets, may be inevitably imperfect. Recent approaches resort to a semi-supervised learning problem by dividing training samples into clean and […]


Continue.. Reliable Label Correction is a Good Booster When Learning with Extremely Noisy Labels

Modeling Ride-Sourcing Matching and Pickup Processes based on Additive Gaussian Process Models

Kavli Affiliate: Zheng Zhu | First 5 Authors: Zheng Zhu, Meng Xu, Yining Di, Xiqun Chen, Jingru Yu | Summary: Matching and pickup processes are core features of ride-sourcing services. Previous studies have adopted abundant analytical models to depict the two processes and obtain operational insights; while the goodness of fit between models and data […]


Continue.. Modeling Ride-Sourcing Matching and Pickup Processes based on Additive Gaussian Process Models

Model predictive control of agro-hydrological systems based on a two-layer neural network modeling framework

Kavli Affiliate: Biao Huang | First 5 Authors: Zhiyinan Huang, Jinfeng Liu, Biao Huang, , | Summary: Water scarcity is an urgent issue to be resolved and improving irrigation water-use efficiency through closed-loop control is essential. The complex agro-hydrological system dynamics, however, often pose challenges in closed-loop control applications. In this work, we propose a […]


Continue.. Model predictive control of agro-hydrological systems based on a two-layer neural network modeling framework

WebFace260M: A Benchmark for Million-Scale Deep Face Recognition

Kavli Affiliate: Zheng Zhu | First 5 Authors: Zheng Zhu, Guan Huang, Jiankang Deng, Yun Ye, Junjie Huang | Summary: Face benchmarks empower the research community to train and evaluate high-performance face recognition systems. In this paper, we contribute a new million-scale recognition benchmark, containing uncurated 4M identities/260M faces (WebFace260M) and cleaned 2M identities/42M faces […]


Continue.. WebFace260M: A Benchmark for Million-Scale Deep Face Recognition

MVSTER: Epipolar Transformer for Efficient Multi-View Stereo

Kavli Affiliate: Zheng Zhu | First 5 Authors: Xiaofeng Wang, Zheng Zhu, Fangbo Qin, Yun Ye, Guan Huang | Summary: Learning-based Multi-View Stereo (MVS) methods warp source images into the reference camera frustum to form 3D volumes, which are fused as a cost volume to be regularized by subsequent networks. The fusing step plays a […]


Continue.. MVSTER: Epipolar Transformer for Efficient Multi-View Stereo

Matrix Entanglement

Kavli Affiliate: Cheng Peng | First 5 Authors: Vaibhav Gautam, Masanori Hanada, Antal Jevicki, Cheng Peng, | Summary: In gauge/gravity duality, matrix degrees of freedom on the gauge theory side play important roles for the emergent geometry. In this paper, we discuss how the entanglement on the gravity side can be described as the entanglement […]


Continue.. Matrix Entanglement

HFT: Lifting Perspective Representations via Hybrid Feature Transformation

Kavli Affiliate: Zheng Zhu | First 5 Authors: Jiayu Zou, Junrui Xiao, Zheng Zhu, Junjie Huang, Guan Huang | Summary: Autonomous driving requires accurate and detailed Bird’s Eye View (BEV) semantic segmentation for decision making, which is one of the most challenging tasks for high-level scene perception. Feature transformation from frontal view to BEV is […]


Continue.. HFT: Lifting Perspective Representations via Hybrid Feature Transformation