检索结果-内蒙古大学图书馆

Adversarial deepfakes: Evaluating vulnerability of deepfake detectors to adversarial examples

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Neekhara, Paarth Hussain, Shehzeen Jere, Malhar Koushanfar, Farinaz McAuley, Julian Department of Computer Science UC San Diego Department of Electrical and Computer Engineering UC San Diego

Recent advances in video manipulation techniques have made the generation of fake videos more accessible than ever before. Manipulated videos can fuel disinformation and reduce trust in media. Therefore detection of fake videos has garnered immense interest in academia and industry. Recently developed Deepfake detection methods rely on deep neural networks (DNNs) to distinguish AI-generated fake videos from real videos. In this work, we demonstrate that it is possible to bypass such detectors by adversarially modifying fake videos synthesized using existing Deepfake generation methods. We further demonstrate that our adversarial perturbations are robust to image and video compression codecs, making them a real-world threat. We present pipelines in both white-box and black-box attack scenarios that can fool DNN based Deepfake detectors into classifying fake videos as real. Copyright © 2020, The Authors. All rights reserved.

关键词： Deep neural networks

FastWave: Accelerating autoregressive convolutional neural networks on FPGA

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Hussain, Shehzeen Javaheripi, Mojan Neekhara, Paarth Kastner, Ryan Koushanfar, Farinaz UC San Diego Department of Electrical and Computer Engineering UC San Diego Department of Computer Science

Autoregressive convolutional neural networks (CNNs) have been widely exploited for sequence generation tasks such as audio synthesis, language modeling and neural machine translation. WaveNet is a deep autoregressive CNN composed of several stacked layers of dilated convolution that is used for sequence generation. While WaveNet produces state-of-the art audio generation results, the naive inference implementation is quite slow;it takes a few minutes to generate just one second of audio on a high-end GPU. In this work, we develop the first accelerator platform FastWave for autoregressive convolutional neural networks, and address the associated design challenges. We design the Fast-Wavenet inference model in Vivado HLS and perform a wide range of optimizations including fixed-point implementation, array partitioning and pipelining. Our model uses a fully parameterized parallel architecture for fast matrix-vector multiplication that enables per-layer customized latency fine-tuning for further throughput improvement. Our experiments comparatively assess the tradeoff between throughput and resource utilization for various optimizations. Our best WaveNet design on the Xilinx XCVU13P FPGA that uses only on-chip memory, achieves 66× faster generation speed compared to CPU implementation and 11× faster generation speed than GPU implementation. Copyright © 2020, The Authors. All rights reserved.

关键词： Field programmable gate arrays (FPGA)

Open-source verification with Chisel and Scala

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Dobis, Andrew Petersen, Tjark Rasmussen, Kasper Juul Hesse Tolotto, Enrico Damsgaard, Hans Jakob Andersen, Simon Thye Lin, Richard Schoeberl, Martin Department of Applied Mathematics and Computer Science Technical University of Denmark Lyngby Denmark Department of Electrical Engineering and Computer Sciences UC Berkeley BerkeleyCA United States

Performance increase with general-purpose processors has come to a halt. We can no longer depend on Moore’s Law to increase computing performance. The only way to achieve higher performance or lower energy consumption is by building domain-specific hardware accelerators. To efficiently design and verify those domain-specific accelerators, we need agile hardware development. One of the main obstacles when proposing such a modern method is the lack of modern tools to attack it. To be able to verify a design in such a time-constrained development method, one needs to have efficient tools both for design and verification. This paper thus proposes ChiselVerify, an open-source tool for verifying circuits described in any Hardware Description Language. It builds on top of the Chisel hardware construction language and uses Scala to drive the verification using a testing strategy inspired by the Universal Verification Methodology (UVM) and adapted for designs described in Chisel. ChiselVerify is created based on three key ideas. First, our solution highly increases the productivity of the verification engineer, by allowing hardware testing to be done in a modern high-level programming environment. Second, the framework functions with any hardware description language thanks to the flexibility of Chisel blackboxes. Finally, the solution is well integrated into the existing Chisel universe, making it an extension of currently existing testing libraries. We implement ChiselVerify in a way inspired by the functionalities found in SystemVerilog. This allows one to use functional coverage, constrained-random verification, bus functional models, transaction-level modeling and much more during the verification process of a design in a contemporary high-level programming ecosystem. © 2021, CC BY-SA.

关键词： General purpose computers

Object Rearrangement Using Learned Implicit Collision Functions

学校读者我要写书评

暂无评论

Object Rearrangement Using Learned Implicit Collision Functi...

IEEE International Conference on Robotics and Automation (ICRA)

作者： Michael Danielczuk Arsalan Mousavian Clemens Eppner Dieter Fox NVIDIA USA UC Berkeley Paul G. Allen School of Computer Science & Engineering University of Washington Seattle WA USA

Robotic object rearrangement combines the skills of picking and placing objects. When object models are unavailable, typical collision-checking models may be unable to predict collisions in partial point clouds with occlusions, making generation of collision-free grasping or placement trajectories challenging. We propose a learned collision model that accepts scene and query object point clouds and predicts collisions for 6DOF object poses within the scene. We train the model on a synthetic set of 1 million scene/object point cloud pairs and 2 billion collision queries. We leverage the learned collision model as part of a model predictive path integral (MPPI) policy in a tabletop rearrangement task and show that the policy can plan collision-free grasps and placements for objects unseen in training in both simulated and physical cluttered scenes with a Franka Panda robot. The learned model outperforms both traditional pipelines and learned ablations by 9.8% in accuracy on a dataset of simulated collision queries and is 75x faster than the best-performing baseline. Videos and supplementary material are available at https://***/publication/2021-03_Object-Rearrangement-Using.

关键词： Training Time-frequency analysis Process control Predictive models Trajectory Planning Task analysis

ANALYTIC DAG CONSTRAINTS FOR DIFFERENTIABLE DAG LEARNING

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Zhang, Zhen Ng, Ignavier Gong, Dong Liu, Yuhang Gong, Mingming Huang, Biwei Zhang, Kun van den Hengel, Anton Shi, Javen Qinfeng Australian Institute for Machine Learning The University of Adelaide Australia Department of Philosophy Carnegie Mellon University United States School of Computer Science and Engineering The University of New South Wales Australia School of Mathematics and Statistics The University of Melbourne Australia Department of Machine Learning Mohamed bin Zayed University of Artificial Intelligence United Arab Emirates UC San Diego United States

Recovering the underlying Directed Acyclic Graph (DAG) structures from observational data presents a formidable challenge, partly due to the combinatorial nature of the DAG-constrained optimization problem. Recently, researchers have identified gradient vanishing as one of the primary obstacles in differentiable DAG learning and have proposed several DAG constraints to mitigate this issue. By developing the necessary theory to establish a connection between analytic functions and DAG constraints, we demonstrate that analytic functions from the set {f(x) = c0 + P∞i=1 cixi|∀i > 0, ci > 0;r = limi→∞ ci/ci+1 > 0} can be employed to formulate effective DAG constraints. Furthermore, we establish that this set of functions is closed under several functional operators, including differentiation, summation, and multiplication. Consequently, these operators can be leveraged to create novel DAG constraints based on existing ones. Using these properties, we design a series of DAG constraints and develop an efficient algorithm to evaluate them. Experiments in various settings demonstrate that our DAG constraints outperform previous state-of-the-art comparators. Our implementation is available at https://***/zzhang1987/AnalyticDAGLearning. © 2025, CC BY-NC-SA.

关键词： Constrained optimization

Building Human Values into Recommender Systems: An Interdisciplinary Synthesis

学校读者我要写书评

暂无评论

ACM Transactions on Recommender Systems 2024年第3期2卷 1-57页

作者： Jonathan Stray Alon Halevy Parisa Assar Dylan Hadfield-Menell Craig Boutilier Amar Ashar Chloe Bakalar Lex Beattie Michael Ekstrand Claire Leibowicz Connie Moon Sehat Sara Johansen Lianne Kerlin David Vickrey Spandana Singh Sanne Vrijenhoek Amy Zhang McKane Andrus Natali Helberger Polina Proutskova Tanushree Mitra Nina Vasan UC Berkeley Center for Human-compatible AI Berkeley USA Meta AI Menlo Park USA Adobe Inc. San Jose USA Department of Electrical Engineering and Computer Science MIT Cambridge USA Google Research Mountain View USA Spotify Inc. New York City USA Meta Inc. Menlo Park USA Department of Information Science Drexel University Philadelphia USA Partnership on AI Palo Alto USA Hacks/Hackers Atlanta USA Department of Psychiatry and Behavioral Sciences Stanford University Stanford USA BBC London UK Institute for Information Law University of Amsterdam Amsterdam Netherlands Allen School of Computer Science and Engineering University of Washington Seattle USA Information School University of Washington Seattle USA Department of Psychiatry and Behavioral Sciences Stanford University Palo Alto USA

Recommender systems are the algorithms which select, filter, and personalize content across many of the world's largest platforms and apps. As such, their positive and negative effects on individuals and on societies have been extensively theorized and studied. Our overarching question is how to ensure that recommender systems enact the values of the individuals and societies that they serve. Addressing this question in a principled fashion requires technical knowledge of recommender design and operation, and also critically depends on insights from diverse fields including social science, ethics, economics, psychology, policy, and law. This article is a multidisciplinary effort to synthesize theory and practice from different perspectives, with the goal of providing a shared language, articulating current design approaches, and identifying open problems. We collect a set of values that seem most relevant to recommender systems operating across different domains, and then examine them from the perspectives of current industry practice, measurement, product design, and policy approaches. Important open problems include multi-stakeholder processes for defining values and resolving trade-offs, better values-driven measurements, recommender controls that people use, non-behavioral algorithmic feedback, optimization for long-term outcomes, causal inference of recommender effects, academic-industry research collaborations, and interdisciplinary policy-making.

关键词： Recommender systems value-sensitive design technology policy

A Fully Programmable On-Chip Planar Waveguide for Machine Learning

学校读者我要写书评

暂无评论

A Fully Programmable On-Chip Planar Waveguide for Machine Le...

Conference on Lasers and Electro-Optics (CLEO)

作者： Martin M. Stein Tatsuhiro Onodera Benjamin A. Ash Mandar M. Sohoni Melissa Bosch Ryotatsu Yanagimoto Marc Jankowski Timothy P. McKenna Tianyu Wang Gennady Shvets Maxim R. Shcherbakov Logan G. Wright Peter L. McMahon School of Applied and Engineering Physics Cornell University Ithaca NY USA NTT Physics and Informatics Laboratories NTT Research Inc. Sunnyvale CA USA E. L. Ginzton Laboratory Stanford University Stanford CA USA Department of Electrical Engineering and Computer Science UC Irvine Irvine CA USA Department of Applied Physics Yale University New Haven CT USA Kavli Institute at Cornell for Nanoscale Science Cornell University Ithaca NY USA

We introduce a device containing a planar waveguide whose spatial refractive index profile $n(x, z)$ can be programmed in real time. We demonstrate use this device as an optical neural network.

ISBN: (纸本)9798350369311

We introduce a device containing a planar waveguide whose spatial refractive index profile $n(x, z)$ can be programmed in real time. We demonstrate use this device as an optical neural network.

关键词： Optical planar waveguides Refractive index Waveguide lasers Optical fiber networks Optical variables control Real-time systems Planar waveguides System-on-chip Optical refraction Optical devices

AWAC: Accelerating online reinforcement learning with offline datasets

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Nair, Ashvin Gupta, Abhishek Dalal, Murtaza Levine, Sergey Department of Electrical Engineering and Computer Science UC Berkeley

Reinforcement learning (RL) provides an appealing formalism for learning control policies from experience. However, the classic active formulation of RL necessitates a lengthy active exploration process for each behavior, making it difficult to apply in real-world settings such as robotic control. If we can instead allow RL algorithms to effectively use previously collected data to aid the online learning process, such applications could be made substantially more practical: the prior data would provide a starting point that mitigates challenges due to exploration and sample complexity, while the online training enables the agent to perfect the desired skill. Such prior data could either constitute expert demonstrations or, more generally, sub-optimal prior data that illustrates potentially useful transitions. While a number of prior methods have either used optimal demonstrations to bootstrap reinforcement learning, or have used sub-optimal data to train purely offline, it remains exceptionally difficult to train a policy with potentially sub-optimal offline data and actually continue to improve it further with online RL. In this paper we systematically analyze why this problem is so challenging, and propose an algorithm that combines sample-efficient dynamic programming with maximum likelihood policy updates, providing a simple and effective framework that is able to leverage large amounts of offline data and then quickly perform online fine-tuning of RL policies. We show that our method, advantage weighted actor critic (AWAC), enables rapid learning of skills with a combination of prior demonstration data and online experience. We demonstrate these benefits on a variety of simulated and real-world robotics domains, including dexterous manipulation with a real multi-fingered hand, drawer opening with a robotic arm, and rotating a valve. Our results show that incorporating prior data can reduce the time required to learn a range of robotic skills to practical time-

关键词： Reinforcement learning

Structural Inversion Asymmetry in Epitaxial Ultrathin Films of Bi (111)/InSb (111)B

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Inbar, Hadass S. Zubair, Muhammad Dong, Jason T. Engel, Aaron N. Dempsey, Connor P. Chang, Yu Hao Nishihaya, Shinichi Khalid, Shoaib Fedorov, Alexei V. Janotti, Anderson Palmstrøm, Chris J. Materials Department UC Santa Barbara CA93106 United States Department of Physics and Astronomy University of Delaware NewarkDE19716 United States Department of Materials Science and Engineering University of Delaware NewarkDE19716 United States Department of Electrical & Computer Engineering UC Santa Barbara CA93106 United States Princeton Plasma Physics Laboratory PrincetonNJ08540 United States Advanced Light Source Lawrence Berkeley National Laboratory BerkeleyCA94720 United States

Bismuth (Bi) films hold potential for spintronic devices due to strong spin-orbit coupling. Understanding the growth, surface states, and interactions with the substrate is key to their functionalization. Large-area high-quality (111) Bi ultrathin films were grown on InSb (111)B substrates by molecular beam epitaxy (MBE). Strong film-substrate interactions epitaxially stabilize the (111) orientation and lead to nonequivalent interface potentials. Analysis of angle-resolved photoemission spectroscopy (ARPES) measurements, employed to characterize the evolution of the surface states with film thickness, indicate a crossing at the M̅ point, suggesting a topologically trivial phase in the thin film. The results show the presence of interfacial bonds to the substrate breaks inversion symmetry, preventing the semimetal-to-semiconductor transition predicted for freestanding bismuth layers, highlighting the importance of controlled functionalization and surface passivation of two-dimensional materials. Copyright © 2023, The Authors. All rights reserved.

关键词： Ultrathin films