15 Hits in 6.4 sec

Score-Based Generative Classifiers [article]

Roland S. Zimmermann, Lukas Schott, Yang Song, Benjamin A. Dunn, David A. Klindt
2021 arXiv   pre-print
We show that these models not only obtain competitive likelihood values but simultaneously achieve state-of-the-art classification accuracy for generative classifiers on CIFAR-10.  ...  Additionally, on natural image datasets, previous results have suggested a trade-off between the likelihood of the data and classification accuracy.  ...  Information Processing Systems 2020, NeurIPS 2020, December 6-12, 2020, virtual, 2020  ... 
arXiv:2110.00473v2 fatcat:6yer6cgkxnbf7kg2m2xmxjtre4

Generalized Likelihood Ratio Test for Adversarially Robust Hypothesis Testing [article]

Bhagyashree Puranik, Upamanyu Madhow, Ramtin Pedarsani
2021 arXiv   pre-print
For non-asymptotic regimes, we show via simulations that the GLRT defense is competitive with the minimax approach under the worst-case attack, while yielding a better robustness-accuracy tradeoff under  ...  , jointly estimating the class of interest and the adversarial perturbation.  ...  tional Conference on Learning Representations, ICLR 2018, Vancouver, on Machine Learning, ICML 2020, 13-18 July 2020, Virtual Event, ser.  ... 
arXiv:2112.02209v1 fatcat:7lh3gixnd5e7zh3b44qnwiyxdy

Fast and Precise: Adjusting Planning Horizon with Adaptive Subgoal Search [article]

Michał Zawalski, Michał Tyrolski, Konrad Czechowski, Damian Stachura, Piotr Piękos, Tomasz Odrzygóźdź, Yuhuai Wu, Łukasz Kuciński, Piotr Miłoś
2022 arXiv   pre-print
In this way, AdaSubS benefits from the efficiency of planning with longer subgoals and the fine control with the shorter ones.  ...  A verification mechanism is employed to filter out unreachable subgoals swiftly and thus allowing to focus on feasible further subgoals.  ...  Information Processing Systems 2020, NeurIPS 2020, December 6-12, 2020, virtual, 2020. 10 [26] Karl Pertsch, Oleh Rybkin, Jingyun Yang, Shenghao Zhou  ... 
arXiv:2206.00702v2 fatcat:fj4x4hljtvcdlm53nxkqcbhvua

Which images to label for few-shot medical landmark detection? [article]

Quan Quan, Qingsong Yao, Jun Li, S. Kevin Zhou
2021 arXiv   pre-print
The advantage of SCP is demonstrated by various experiments on three widely-used public datasets.  ...  Few-shot learning is developed to alleviate this burden, which achieves competitive performances with only several labeled data.  ...  In the ference on Neural Information Processing Systems 2020, future, we plan to further improve the efficiency and ac- NeurIPS 2020, December 6-12, 2020, virtual, 2020  ... 
arXiv:2112.04386v2 fatcat:ajn62lhsz5bl7hruzmujcntc3u

Neural Network Module Decomposition and Recomposition [article]

Hiroaki Kingetsu, Kenichi Kobayashi, Taiji Suzuki
2021 arXiv   pre-print
We demonstrate that the proposed method can decompose and recompose DNNs with high compression ratio and high accuracy and is superior to the existing method through sharing weights between modules.  ...  Decomposed modules are expected to have the advantages of interpretability and verifiability due to their small size.  ...  In Advances in Neural Information Processing Systems 32: Annual Conference on Neural Information Processing Sys- tems 2019, NeurIPS 2019, December 8-14, 2019, Vancouver, BC, Canada, 3592–3602.  ... 
arXiv:2112.13208v1 fatcat:zcccxh6nmrberacsaj3b5ct4vi

High Fidelity Visualization of What Your Self-Supervised Representation Knows About [article]

Florian Bordes, Randall Balestriero, Pascal Vincent
2021 arXiv   pre-print
We further demonstrate how this model's generation quality is on par with state-of-the-art generative models while being faithful to the representation used as conditioning.  ...  Processing Systems 2020, NeurIPS 2020, December 6-12, 2020, virtual, 2020.  ...  Processing Systems 2020, NeurIPS 2020, December 6-12, 2020, virtual, 2020.  ... 
arXiv:2112.09164v1 fatcat:6lk3n7w3crbgpomjdjanm7xhju

TeachMyAgent: a Benchmark for Automatic Curriculum Learning in Deep RL [article]

Clément Romac, Rémy Portelas, Katja Hofmann, Pierre-Yves Oudeyer
2021 arXiv   pre-print
We then use TeachMyAgent to conduct a comparative study of representative existing approaches, showcasing the competitiveness of some ACL algorithms that do not use expert knowledge.  ...  It includes 1) challenge-specific unit-tests using variants of a procedural Box2D bipedal walker environment, and 2) a new procedural Parkour environment combining most ACL challenges, making it ideal  ...  .), Advances in Neural Information Processing Systems 33: Annual Conference on Neural Information Processing Systems 2020, NeurIPS 2020, December 6-12, 2020, virtual, 2020.  ... 
arXiv:2103.09815v2 fatcat:qtsp6ghsgrd47fwgznnqdwrtsu

What to Pre-Train on? Efficient Intermediate Task Selection [article]

Clifton Poth, Jonas Pfeiffer, Andreas Rücklé, Iryna Gurevych
2021 arXiv   pre-print
We experiment with a diverse set of 42 intermediate and 11 target English classification, multiple choice, question answering, and sequence tagging tasks.  ...  Our best methods achieve an average Regret@3 of less than 1% across all target tasks, demonstrating that we are able to efficiently identify the best datasets for intermediate training.  ...  We thank Leonardo Ribeiro and the anonymous reviewers for insightful feedback and suggestions on a draft of this paper.  ... 
arXiv:2104.08247v2 fatcat:4ljcfshev5f3tmgugrrrkh3s4m

Evaluation of Socially-Aware Robot Navigation

Yuxiang Gao, Chien-Ming Huang
2022 Frontiers in Robotics and AI  
As mobile robots are increasingly introduced into our daily lives, it grows ever more imperative that these robots navigate with and among people in a safe and socially acceptable manner, particularly  ...  As an effort to aid more productive development and progress comparisons, in this paper we review the evaluation methods, scenarios, datasets, and metrics commonly used in previous socially-aware navigation  ...  “SocioSense: Robot Navigation Amongst Pedestrians with Social and Psychological Constraints,” in IEEE International Conference on Intelligent Robots and Systems, Vancouver, BC, Canada, 24-28 Sept. 2017  ... 
doi:10.3389/frobt.2021.721317 pmid:35096978 pmcid:PMC8791647 fatcat:egddc7tddbhj7hyevenemyh6pi

Moreau-Yosida f-divergences [article]

Dávid Terjék
2021 arXiv   pre-print
total variation divergences as GANs trained on CIFAR-10, leading to competitive results and a simple solution to the problem of uniqueness of the optimal critic.  ...  The corresponding variational formulas provide a generalization of a number of recent results, novel special cases of interest and a relaxation of the hard Lipschitz constraint.  ...  .), Advances in Neural Information Processing Systems 32: Annual Conference on Neural Information Processing Systems 2019, NeurIPS 2019, December 8-14, 2019, Vancouver, BC, Canada, pp. 413-422, 2019.  ... 
arXiv:2102.13416v2 fatcat:qn3y27yxirdihbeulitinlbkfe

Double Sampling Randomized Smoothing [article]

Linyi Li and Jiawei Zhang and Tao Xie and Bo Li
2022 arXiv   pre-print
We instantiate DSRS for a generalized family of Gaussian smoothing and propose an efficient and sound computing method based on customized dual optimization considering sampling error.  ...  Extensive experiments on MNIST, CIFAR-10, and ImageNet verify our theory and show that DSRS certifies larger robust radii than existing baselines consistently under different settings.  ...  This work is partially supported by NSF grant No.1910100, NSF CNS No.2046726, C3 AI, and the Alfred P. Sloan Foundation.  ... 
arXiv:2206.07912v2 fatcat:axqzuupezjfulgdw2bojlexqoa

Jigsaw: Large Language Models meet Program Synthesis [article]

Naman Jain, Skanda Vaidyanath, Arun Iyer, Nagarajan Natarajan, Suresh Parthasarathy, Sriram Rajamani, Rahul Sharma
2021 arXiv   pre-print
We view these developments with a mixture of optimism and caution.  ...  In this paper, we present an approach to augment these large language models with post-processing steps based on program analysis and synthesis techniques, that understand the syntax and semantics of programs  ...  , NeurIPS 2020, December 6-12, 2020, 8 CONCLUSION AND FUTURE WORK virtual, Hugo Larochelle, Marc’Aurelio Ranzato, Raia Hadsell, Maria-Florina  ... 
arXiv:2112.02969v1 fatcat:kjdac665kzhthldiqckru3in3u

Towards Explainable Fact Checking [article]

Isabelle Augenstein
2021 arXiv   pre-print
, and, very recently, by legislation requiring online platforms operating in the EU to provide transparent reporting on their services.  ...  This thesis presents my research on automatic fact checking, including claim check-worthiness detection, stance detection and veracity prediction.  ...  In: The 3rd NeurIPS Workshop on Conversational AI. Vancouver, Canada. url: http://alborz- Conversational-AI/Papers/6.pdf 30.  ... 
arXiv:2108.10274v2 fatcat:5s4an6irezcjfmvvhmiaeqarh4

PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization [article]

Jingqing Zhang, Yao Zhao, Mohammad Saleh, Peter J. Liu
2020 arXiv   pre-print
Experiments demonstrate it achieves state-of-the-art performance on all 12 downstream datasets measured by ROUGE scores.  ...  We evaluated our best PEGASUS model on 12 downstream summarization tasks spanning news, science, stories, instructions, emails, patents, and legislative bills.  ...  , and Samy Bengio for reviewing the manuscript.  ... 
arXiv:1912.08777v3 fatcat:sniaydtda5a35pgyugitvcjtpy

Διερεύνηση της γεωλογικής δομής και της τεκτονικής παραμόρφωσης της Σερβομακεδονικής Μάζας στα όρη των Κερδυλλίων και Βερτίσκου (Βόρεια Ελλάδα)

Αναστάσιος Π. Πλούγαρλης
Six deformational events (D1-D6) were i [...]  ...  The study area is located in northern Greece and occupies the Vertiskos and Kerdyllion Mountains.  ...  In Advances in Neural Information Processing Systems 32: Annual Conference on Neural Information Processing Systems 2019, NeurIPS 2019, 8-14 December 2019, Vancouver, BC, Canada, pages 5754–5764.  ... 
doi:10.26262/ fatcat:snneggenqzdnvbl7kn7zpztqqq