October 20, 2019

3048 words 15 mins read

Paper Group AWR 192

Backpropagation for Implicit Spectral Densities. Pseudo-Bayesian Learning with Kernel Fourier Transform as Prior. ML-Leaks: Model and Data Independent Membership Inference Attacks and Defenses on Machine Learning Models. HoloFace: Augmenting Human-to-Human Interactions on HoloLens. CBAM: Convolutional Block Attention Module. IGCV3: Interleaved Low- …

Backpropagation for Implicit Spectral Densities


Title	Backpropagation for Implicit Spectral Densities
Authors	Aditya Ramesh, Yann LeCun
Abstract	Most successful machine intelligence systems rely on gradient-based learning, which is made possible by backpropagation. Some systems are designed to aid us in interpreting data when explicit goals cannot be provided. These unsupervised systems are commonly trained by backpropagating through a likelihood function. We introduce a tool that allows us to do this even when the likelihood is not explicitly set, by instead using the implicit likelihood of the model. Explicitly defining the likelihood often entails making heavy-handed assumptions that impede our ability to solve challenging tasks. On the other hand, the implicit likelihood of the model is accessible without the need for such assumptions. Our tool, which we call spectral backpropagation, allows us to optimize it in much greater generality than what has been attempted before. GANs can also be viewed as a technique for optimizing implicit likelihoods. We study them using spectral backpropagation in order to demonstrate robustness for high-dimensional problems, and identify two novel properties of the generator G: (1) there exist aberrant, nonsensical outputs to which G assigns very high likelihood, and (2) the eigenvectors of the metric induced by G over latent space correspond to quasi-disentangled explanatory factors.
Tasks
Published	2018-06-01
URL	http://arxiv.org/abs/1806.00499v1
PDF	http://arxiv.org/pdf/1806.00499v1.pdf
PWC	https://paperswithcode.com/paper/backpropagation-for-implicit-spectral
Repo	https://github.com/EiffL/SpectralFlow
Framework	tf

Pseudo-Bayesian Learning with Kernel Fourier Transform as Prior


Title	Pseudo-Bayesian Learning with Kernel Fourier Transform as Prior
Authors	Gaël Letarte, Emilie Morvant, Pascal Germain
Abstract	We revisit Rahimi and Recht (2007)‘s kernel random Fourier features (RFF) method through the lens of the PAC-Bayesian theory. While the primary goal of RFF is to approximate a kernel, we look at the Fourier transform as a prior distribution over trigonometric hypotheses. It naturally suggests learning a posterior on these hypotheses. We derive generalization bounds that are optimized by learning a pseudo-posterior obtained from a closed-form expression. Based on this study, we consider two learning strategies: The first one finds a compact landmarks-based representation of the data where each landmark is given by a distribution-tailored similarity measure, while the second one provides a PAC-Bayesian justification to the kernel alignment method of Sinha and Duchi (2016).
Tasks
Published	2018-10-30
URL	http://arxiv.org/abs/1810.12683v2
PDF	http://arxiv.org/pdf/1810.12683v2.pdf
PWC	https://paperswithcode.com/paper/pseudo-bayesian-learning-with-kernel-fourier
Repo	https://github.com/gletarte/pbrff
Framework	none

ML-Leaks: Model and Data Independent Membership Inference Attacks and Defenses on Machine Learning Models


Title	ML-Leaks: Model and Data Independent Membership Inference Attacks and Defenses on Machine Learning Models
Authors	Ahmed Salem, Yang Zhang, Mathias Humbert, Pascal Berrang, Mario Fritz, Michael Backes
Abstract	Machine learning (ML) has become a core component of many real-world applications and training data is a key factor that drives current progress. This huge success has led Internet companies to deploy machine learning as a service (MLaaS). Recently, the first membership inference attack has shown that extraction of information on the training set is possible in such MLaaS settings, which has severe security and privacy implications. However, the early demonstrations of the feasibility of such attacks have many assumptions on the adversary, such as using multiple so-called shadow models, knowledge of the target model structure, and having a dataset from the same distribution as the target model’s training data. We relax all these key assumptions, thereby showing that such attacks are very broadly applicable at low cost and thereby pose a more severe risk than previously thought. We present the most comprehensive study so far on this emerging and developing threat using eight diverse datasets which show the viability of the proposed attacks across domains. In addition, we propose the first effective defense mechanisms against such broader class of membership inference attacks that maintain a high level of utility of the ML model.
Tasks	Inference Attack
Published	2018-06-04
URL	http://arxiv.org/abs/1806.01246v2
PDF	http://arxiv.org/pdf/1806.01246v2.pdf
PWC	https://paperswithcode.com/paper/ml-leaks-model-and-data-independent
Repo	https://github.com/AhmedSalem2/ML-Leaks
Framework	none

HoloFace: Augmenting Human-to-Human Interactions on HoloLens


Title	HoloFace: Augmenting Human-to-Human Interactions on HoloLens
Authors	Marek Kowalski, Zbigniew Nasarzewski, Grzegorz Galinski, Piotr Garbat
Abstract	We present HoloFace, an open-source framework for face alignment, head pose estimation and facial attribute retrieval for Microsoft HoloLens. HoloFace implements two state-of-the-art face alignment methods which can be used interchangeably: one running locally and one running on a remote backend. Head pose estimation is accomplished by fitting a deformable 3D model to the landmarks localized using face alignment. The head pose provides both the rotation of the head and a position in the world space. The parameters of the fitted 3D face model provide estimates of facial attributes such as mouth opening or smile. Together the above information can be used to augment the faces of people seen by the HoloLens user, and thus their interaction. Potential usage scenarios include facial recognition, emotion recognition, eye gaze tracking and many others. We demonstrate the capabilities of our framework by augmenting the faces of people seen through the HoloLens with various objects and animations.
Tasks	Emotion Recognition, Face Alignment, Head Pose Estimation, Pose Estimation
Published	2018-02-01
URL	http://arxiv.org/abs/1802.00278v1
PDF	http://arxiv.org/pdf/1802.00278v1.pdf
PWC	https://paperswithcode.com/paper/holoface-augmenting-human-to-human
Repo	https://github.com/MarekKowalski/HoloFace
Framework	none

CBAM: Convolutional Block Attention Module


Title	CBAM: Convolutional Block Attention Module
Authors	Sanghyun Woo, Jongchan Park, Joon-Young Lee, In So Kweon
Abstract	We propose Convolutional Block Attention Module (CBAM), a simple yet effective attention module for feed-forward convolutional neural networks. Given an intermediate feature map, our module sequentially infers attention maps along two separate dimensions, channel and spatial, then the attention maps are multiplied to the input feature map for adaptive feature refinement. Because CBAM is a lightweight and general module, it can be integrated into any CNN architectures seamlessly with negligible overheads and is end-to-end trainable along with base CNNs. We validate our CBAM through extensive experiments on ImageNet-1K, MS~COCO detection, and VOC~2007 detection datasets. Our experiments show consistent improvements in classification and detection performances with various models, demonstrating the wide applicability of CBAM. The code and models will be publicly available.
Tasks	Image Classification
Published	2018-07-17
URL	http://arxiv.org/abs/1807.06521v2
PDF	http://arxiv.org/pdf/1807.06521v2.pdf
PWC	https://paperswithcode.com/paper/cbam-convolutional-block-attention-module
Repo	https://github.com/vinthony/s2am
Framework	pytorch

IGCV3: Interleaved Low-Rank Group Convolutions for Efficient Deep Neural Networks


Title	IGCV3: Interleaved Low-Rank Group Convolutions for Efficient Deep Neural Networks
Authors	Ke Sun, Mingjie Li, Dong Liu, Jingdong Wang
Abstract	In this paper, we are interested in building lightweight and efficient convolutional neural networks. Inspired by the success of two design patterns, composition of structured sparse kernels, e.g., interleaved group convolutions (IGC), and composition of low-rank kernels, e.g., bottle-neck modules, we study the combination of such two design patterns, using the composition of structured sparse low-rank kernels, to form a convolutional kernel. Rather than introducing a complementary condition over channels, we introduce a loose complementary condition, which is formulated by imposing the complementary condition over super-channels, to guide the design for generating a dense convolutional kernel. The resulting network is called IGCV3. We empirically demonstrate that the combination of low-rank and sparse kernels boosts the performance and the superiority of our proposed approach to the state-of-the-arts, IGCV2 and MobileNetV2 over image classification on CIFAR and ImageNet and object detection on COCO.
Tasks	Image Classification, Object Detection
Published	2018-06-01
URL	http://arxiv.org/abs/1806.00178v2
PDF	http://arxiv.org/pdf/1806.00178v2.pdf
PWC	https://paperswithcode.com/paper/igcv3-interleaved-low-rank-group-convolutions
Repo	https://github.com/homles11/IGCV3
Framework	tf

Practical Full Resolution Learned Lossless Image Compression


Title	Practical Full Resolution Learned Lossless Image Compression
Authors	Fabian Mentzer, Eirikur Agustsson, Michael Tschannen, Radu Timofte, Luc Van Gool
Abstract	We propose the first practical learned lossless image compression system, L3C, and show that it outperforms the popular engineered codecs, PNG, WebP and JPEG 2000. At the core of our method is a fully parallelizable hierarchical probabilistic model for adaptive entropy coding which is optimized end-to-end for the compression task. In contrast to recent autoregressive discrete probabilistic models such as PixelCNN, our method i) models the image distribution jointly with learned auxiliary representations instead of exclusively modeling the image distribution in RGB space, and ii) only requires three forward-passes to predict all pixel probabilities instead of one for each pixel. As a result, L3C obtains over two orders of magnitude speedups when sampling compared to the fastest PixelCNN variant (Multiscale-PixelCNN). Furthermore, we find that learning the auxiliary representation is crucial and outperforms predefined auxiliary representations such as an RGB pyramid significantly.
Tasks	Image Compression
Published	2018-11-30
URL	https://arxiv.org/abs/1811.12817v3
PDF	https://arxiv.org/pdf/1811.12817v3.pdf
PWC	https://paperswithcode.com/paper/practical-full-resolution-learned-lossless
Repo	https://github.com/fab-jul/L3C-PyTorch
Framework	pytorch

Generating 3D Adversarial Point Clouds


Title	Generating 3D Adversarial Point Clouds
Authors	Chong Xiang, Charles R. Qi, Bo Li
Abstract	Deep neural networks are known to be vulnerable to adversarial examples which are carefully crafted instances to cause the models to make wrong predictions. While adversarial examples for 2D images and CNNs have been extensively studied, less attention has been paid to 3D data such as point clouds. Given many safety-critical 3D applications such as autonomous driving, it is important to study how adversarial point clouds could affect current deep 3D models. In this work, we propose several novel algorithms to craft adversarial point clouds against PointNet, a widely used deep neural network for point cloud processing. Our algorithms work in two ways: adversarial point perturbation and adversarial point generation. For point perturbation, we shift existing points negligibly. For point generation, we generate either a set of independent and scattered points or a small number (1-3) of point clusters with meaningful shapes such as balls and airplanes which could be hidden in the human psyche. In addition, we formulate six perturbation measurement metrics tailored to the attacks in point clouds and conduct extensive experiments to evaluate the proposed algorithms on the ModelNet40 3D shape classification dataset. Overall, our attack algorithms achieve a success rate higher than 99% for all targeted attacks
Tasks	Autonomous Driving
Published	2018-09-19
URL	https://arxiv.org/abs/1809.07016v4
PDF	https://arxiv.org/pdf/1809.07016v4.pdf
PWC	https://paperswithcode.com/paper/generating-3d-adversarial-point-clouds
Repo	https://github.com/xiangchong1/3d-adv-pc
Framework	tf

Positive-unlabeled convolutional neural networks for particle picking in cryo-electron micrographs


Title	Positive-unlabeled convolutional neural networks for particle picking in cryo-electron micrographs
Authors	Tristan Bepler, Andrew Morin, Julia Brasch, Lawrence Shapiro, Alex J. Noble, Bonnie Berger
Abstract	Cryo-electron microscopy (cryoEM) is an increasingly popular method for protein structure determination. However, identifying a sufficient number of particles for analysis (often >100,000) can take months of manual effort. Current computational approaches are limited by high false positive rates and require significant ad-hoc post-processing, especially for unusually shaped particles. To address this shortcoming, we develop Topaz, an efficient and accurate particle picking pipeline using neural networks trained with few labeled particles by newly leveraging the remaining unlabeled particles through the framework of positive-unlabeled (PU) learning. Remarkably, despite using minimal labeled particles, Topaz allows us to improve reconstruction resolution by up to 0.15 {\AA} over published particles on three public cryoEM datasets without any post-processing. Furthermore, we show that our novel generalized-expectation criteria approach to PU learning outperforms existing general PU learning approaches when applied to particle detection, especially for challenging datasets of non-globular proteins. We expect Topaz to be an essential component of cryoEM analysis.
Tasks
Published	2018-03-22
URL	http://arxiv.org/abs/1803.08207v2
PDF	http://arxiv.org/pdf/1803.08207v2.pdf
PWC	https://paperswithcode.com/paper/positive-unlabeled-convolutional-neural
Repo	https://github.com/tbepler/topaz
Framework	pytorch

Automatic differentiation in ML: Where we are and where we should be going


Title	Automatic differentiation in ML: Where we are and where we should be going
Authors	Bart van Merriënboer, Olivier Breuleux, Arnaud Bergeron, Pascal Lamblin
Abstract	We review the current state of automatic differentiation (AD) for array programming in machine learning (ML), including the different approaches such as operator overloading (OO) and source transformation (ST) used for AD, graph-based intermediate representations for programs, and source languages. Based on these insights, we introduce a new graph-based intermediate representation (IR) which specifically aims to efficiently support fully-general AD for array programming. Unlike existing dataflow programming representations in ML frameworks, our IR naturally supports function calls, higher-order functions and recursion, making ML models easier to implement. The ability to represent closures allows us to perform AD using ST without a tape, making the resulting derivative (adjoint) program amenable to ahead-of-time optimization using tools from functional language compilers, and enabling higher-order derivatives. Lastly, we introduce a proof of concept compiler toolchain called Myia which uses a subset of Python as a front end.
Tasks
Published	2018-10-26
URL	http://arxiv.org/abs/1810.11530v2
PDF	http://arxiv.org/pdf/1810.11530v2.pdf
PWC	https://paperswithcode.com/paper/automatic-differentiation-in-ml-where-we-are
Repo	https://github.com/mila-udem/myia
Framework	pytorch

Segmentation of Drosophila Heart in Optical Coherence Microscopy Images Using Convolutional Neural Networks


Title	Segmentation of Drosophila Heart in Optical Coherence Microscopy Images Using Convolutional Neural Networks
Authors	Lian Duan, Xi Qin, Yuanhao He, Xialin Sang, Jinda Pan, Tao Xu, Jing Men, Rudolph E. Tanzi, Airong Li, Yutao Ma, Chao Zhou
Abstract	Convolutional neural networks are powerful tools for image segmentation and classification. Here, we use this method to identify and mark the heart region of Drosophila at different developmental stages in the cross-sectional images acquired by a custom optical coherence microscopy (OCM) system. With our well-trained convolutional neural network model, the heart regions through multiple heartbeat cycles can be marked with an intersection over union (IOU) of ~86%. Various morphological and dynamical cardiac parameters can be quantified accurately with automatically segmented heart regions. This study demonstrates an efficient heart segmentation method to analyze OCM images of the beating heart in Drosophila.
Tasks	Semantic Segmentation
Published	2018-03-05
URL	http://arxiv.org/abs/1803.01947v2
PDF	http://arxiv.org/pdf/1803.01947v2.pdf
PWC	https://paperswithcode.com/paper/segmentation-of-drosophila-heart-in-optical
Repo	https://github.com/Nogimon/FlyNet
Framework	tf

NTUA-SLP at SemEval-2018 Task 1: Predicting Affective Content in Tweets with Deep Attentive RNNs and Transfer Learning


Title	NTUA-SLP at SemEval-2018 Task 1: Predicting Affective Content in Tweets with Deep Attentive RNNs and Transfer Learning
Authors	Christos Baziotis, Nikos Athanasiou, Alexandra Chronopoulou, Athanasia Kolovou, Georgios Paraskevopoulos, Nikolaos Ellinas, Shrikanth Narayanan, Alexandros Potamianos
Abstract	In this paper we present deep-learning models that submitted to the SemEval-2018 Task~1 competition: “Affect in Tweets”. We participated in all subtasks for English tweets. We propose a Bi-LSTM architecture equipped with a multi-layer self attention mechanism. The attention mechanism improves the model performance and allows us to identify salient words in tweets, as well as gain insight into the models making them more interpretable. Our model utilizes a set of word2vec word embeddings trained on a large collection of 550 million Twitter messages, augmented by a set of word affective features. Due to the limited amount of task-specific training data, we opted for a transfer learning approach by pretraining the Bi-LSTMs on the dataset of Semeval 2017, Task 4A. The proposed approach ranked 1st in Subtask E “Multi-Label Emotion Classification”, 2nd in Subtask A “Emotion Intensity Regression” and achieved competitive results in other subtasks.
Tasks	Emotion Classification, Transfer Learning, Word Embeddings
Published	2018-04-18
URL	http://arxiv.org/abs/1804.06658v1
PDF	http://arxiv.org/pdf/1804.06658v1.pdf
PWC	https://paperswithcode.com/paper/ntua-slp-at-semeval-2018-task-1-predicting
Repo	https://github.com/cbaziotis/ntua-slp-semeval2018
Framework	pytorch

NTUA-SLP at IEST 2018: Ensemble of Neural Transfer Methods for Implicit Emotion Classification


Title	NTUA-SLP at IEST 2018: Ensemble of Neural Transfer Methods for Implicit Emotion Classification
Authors	Alexandra Chronopoulou, Aikaterini Margatina, Christos Baziotis, Alexandros Potamianos
Abstract	In this paper we present our approach to tackle the Implicit Emotion Shared Task (IEST) organized as part of WASSA 2018 at EMNLP 2018. Given a tweet, from which a certain word has been removed, we are asked to predict the emotion of the missing word. In this work, we experiment with neural Transfer Learning (TL) methods. Our models are based on LSTM networks, augmented with a self-attention mechanism. We use the weights of various pretrained models, for initializing specific layers of our networks. We leverage a big collection of unlabeled Twitter messages, for pretraining word2vec word embeddings and a set of diverse language models. Moreover, we utilize a sentiment analysis dataset for pretraining a model, which encodes emotion related information. The submitted model consists of an ensemble of the aforementioned TL models. Our team ranked 3rd out of 30 participants, achieving an F1 score of 0.703.
Tasks	Emotion Classification, Sentiment Analysis, Transfer Learning, Word Embeddings
Published	2018-09-03
URL	http://arxiv.org/abs/1809.00717v1
PDF	http://arxiv.org/pdf/1809.00717v1.pdf
PWC	https://paperswithcode.com/paper/ntua-slp-at-iest-2018-ensemble-of-neural
Repo	https://github.com/alexandra-chron/wassa-2018
Framework	pytorch

Universal Planning Networks


Title	Universal Planning Networks
Authors	Aravind Srinivas, Allan Jabri, Pieter Abbeel, Sergey Levine, Chelsea Finn
Abstract	A key challenge in complex visuomotor control is learning abstract representations that are effective for specifying goals, planning, and generalization. To this end, we introduce universal planning networks (UPN). UPNs embed differentiable planning within a goal-directed policy. This planning computation unrolls a forward model in a latent space and infers an optimal action plan through gradient descent trajectory optimization. The plan-by-gradient-descent process and its underlying representations are learned end-to-end to directly optimize a supervised imitation learning objective. We find that the representations learned are not only effective for goal-directed visual imitation via gradient-based trajectory optimization, but can also provide a metric for specifying goals using images. The learned representations can be leveraged to specify distance-based rewards to reach new target states for model-free reinforcement learning, resulting in substantially more effective learning when solving new tasks described via image-based goals. We were able to achieve successful transfer of visuomotor planning strategies across robots with significantly different morphologies and actuation capabilities.
Tasks	Imitation Learning
Published	2018-04-02
URL	http://arxiv.org/abs/1804.00645v2
PDF	http://arxiv.org/pdf/1804.00645v2.pdf
PWC	https://paperswithcode.com/paper/universal-planning-networks
Repo	https://github.com/aravindsrinivas/upn
Framework	tf

ChauffeurNet: Learning to Drive by Imitating the Best and Synthesizing the Worst


Title	ChauffeurNet: Learning to Drive by Imitating the Best and Synthesizing the Worst
Authors	Mayank Bansal, Alex Krizhevsky, Abhijit Ogale
Abstract	Our goal is to train a policy for autonomous driving via imitation learning that is robust enough to drive a real vehicle. We find that standard behavior cloning is insufficient for handling complex driving scenarios, even when we leverage a perception system for preprocessing the input and a controller for executing the output on the car: 30 million examples are still not enough. We propose exposing the learner to synthesized data in the form of perturbations to the expert’s driving, which creates interesting situations such as collisions and/or going off the road. Rather than purely imitating all data, we augment the imitation loss with additional losses that penalize undesirable events and encourage progress – the perturbations then provide an important signal for these losses and lead to robustness of the learned model. We show that the ChauffeurNet model can handle complex situations in simulation, and present ablation experiments that emphasize the importance of each of our proposed changes and show that the model is responding to the appropriate causal factors. Finally, we demonstrate the model driving a car in the real world.
Tasks	Autonomous Driving, Imitation Learning
Published	2018-12-07
URL	http://arxiv.org/abs/1812.03079v1
PDF	http://arxiv.org/pdf/1812.03079v1.pdf
PWC	https://paperswithcode.com/paper/chauffeurnet-learning-to-drive-by-imitating
Repo	https://github.com/kdhingra307/ChauffeurNet
Framework	none