More Like this

1. Advances in neural information processing systems

   Lyu, H; Sha, N.; Qin, S.; Yan, M.; Xie, Y.; Wang, R. (December 2019, Advances in neural information processing systems)

   This paper extends robust principal component analysis (RPCA) to nonlinear manifolds. Suppose that the observed data matrix is the sum of a sparse component and a component drawn from some low dimensional manifold. Is it possible to separate them by using similar ideas as RPCA? Is there any benefit in treating the manifold as a whole as opposed to treating each local region independently? We answer these two questions affirmatively by proposing and analyzing an optimization framework that separates the sparse component from the manifold under noisy data. Theoretical error bounds are provided when the tangent spaces of the manifold satisfy certain incoherence conditions. We also provide a near optimal choice of the tuning parameters for the proposed optimization formulation with the help of a new curvature estimation method. The efficacy of our method is demonstrated on both synthetic and real datasets. 

   more » « less
2. Neural Information Processing

   https://doi.org/10.1002/9780470015902.a0029420  

   Gabbiani, Fabrizio; Midtgaard, Jens (January 2022, Encyclopedia of life sciences)

   The central nervous system (CNS) is specialised in processing information originating from a variety of internal and external sources. Its basic information processing units are nerve cells or neu- rons. Within the CNS, these elementary building blocks are densely interconnected in hierarchical and parallel pathways. Information originating from sensory neurons in contact with the body periphery is gradually transformed along these pathways to generate specific actions through signals relayed by motor neurons to peripheral organs. Information processing in the CNS exhibits an astonishing degree of diversity when con- sidering the morphology of its neurons, their connectivity patterns, their electrical and their biochemical signalling properties. Thus, the infor- mation processing capabilities of animal brains are quite distinct from those of existing man-made machines, being characterised by their resilience to noise, their capacity to rapidly adapt, learn and generalise, as well as their ability to participate in complex social behaviours. 

   more » « less
3. Batched Thompson Sampling, Advances in Neural Information Processing Systems

   Kalkanli, Cem and (December 2021, Advances in neural information processing systems)
4. Multi-source Domain Adaptation for Semantic Segmentation”, Advances in Neural Information Processing Systems

   Zhao, S.; Li, B.; Yue, X.; Gu, Y.; Xu, P.; Hu, R.; Chai, H.; Keutzer, K. (October 2019, Advances in neural information processing systems)

   Simulation-to-real domain adaptation for semantic segmentation has been actively studied for various applications such as autonomous driving. Existing methods mainly focus on a single-source setting, which cannot easily handle a more practical scenario of multiple sources with different distributions. In this paper, we propose to investigate multi-source domain adaptation for semantic segmentation. Specifically, we design a novel framework, termed Multi-source Adversarial Domain Aggregation Network (MADAN), which can be trained in an end-to-end manner. First, we generate an adapted domain for each source with dynamic semantic consistency while aligning at the pixel-level cycle-consistently towards the target. Second, we propose sub-domain aggregation discriminator and cross-domain cycle discriminator to make different adapted domains more closely aggregated. Finally, feature-level alignment is performed between the aggregated domain and target domain while training the segmentation network. Extensive experiments from synthetic GTA and SYNTHIA to real Cityscapes and BDDS datasets demonstrate that the proposed MADAN model outperforms state-of-the-art approaches. Our source code is released at: https://github.com/Luodian/MADAN. 

   more » « less
5. Estimating the Rate-Distortion Function by Wasserstein Gradient Descent 37th Conference on Neural Information Processing Systems (NeurIPS)

   Yang, Yibo; Eckstein, Stephan; Nutz, Marcel; Mandt, Stephan (November 2023, 37th Conference on Neural Information Processing Systems (NeurIPS))

   A. Oh; T. Naumann; A. Globerson; K. Saenko; M. Hardt; S. Levine (Ed.)

   In the theory of lossy compression, the rate-distortion (R-D) function R(D) describes how much a data source can be compressed (in bit-rate) at any given level of fidelity (distortion). Obtaining R(D) for a given data source establishes the fundamental performance limit for all compression algorithms. We propose a new method to estimate R(D) from the perspective of optimal transport. Unlike the classic Blahut--Arimoto algorithm which fixes the support of the reproduction distribution in advance, our Wasserstein gradient descent algorithm learns the support of the optimal reproduction distribution by moving particles. We prove its local convergence and analyze the sample complexity of our R-D estimator based on a connection to entropic optimal transport. Experimentally, we obtain comparable or tighter bounds than state-of-the-art neural network methods on low-rate sources while requiring considerably less tuning and computation effort. We also highlight a connection to maximum-likelihood deconvolution and introduce a new class of sources that can be used as test cases with known solutions to the R-D problem. 

   more » « less