Paper Group AWR 78
Referential Uncertainty and Word Learning in High-dimensional, Continuous Meaning Spaces. Clockwork Convnets for Video Semantic Segmentation. RL$^2$: Fast Reinforcement Learning via Slow Reinforcement Learning. Instance Normalization: The Missing Ingredient for Fast Stylization. A Baseline for Detecting Misclassified and Out-of-Distribution Example …
Referential Uncertainty and Word Learning in High-dimensional, Continuous Meaning Spaces
| Title | Referential Uncertainty and Word Learning in High-dimensional, Continuous Meaning Spaces |
| Authors | Michael Spranger, Katrien Beuls |
| Abstract | This paper discusses lexicon word learning in high-dimensional meaning spaces from the viewpoint of referential uncertainty. We investigate various state-of-the-art Machine Learning algorithms and discuss the impact of scaling, representation and meaning space structure. We demonstrate that current Machine Learning techniques successfully deal with high-dimensional meaning spaces. In particular, we show that exponentially increasing dimensions linearly impact learner performance and that referential uncertainty from word sensitivity has no impact. |
| Tasks | |
| Published | 2016-09-30 |
| URL | http://arxiv.org/abs/1609.09580v1 |
| http://arxiv.org/pdf/1609.09580v1.pdf | |
| PWC | https://paperswithcode.com/paper/referential-uncertainty-and-word-learning-in |
| Repo | https://github.com/mspranger/icdl2016language |
| Framework | none |
Clockwork Convnets for Video Semantic Segmentation
| Title | Clockwork Convnets for Video Semantic Segmentation |
| Authors | Evan Shelhamer, Kate Rakelly, Judy Hoffman, Trevor Darrell |
| Abstract | Recent years have seen tremendous progress in still-image segmentation; however the na"ive application of these state-of-the-art algorithms to every video frame requires considerable computation and ignores the temporal continuity inherent in video. We propose a video recognition framework that relies on two key observations: 1) while pixels may change rapidly from frame to frame, the semantic content of a scene evolves more slowly, and 2) execution can be viewed as an aspect of architecture, yielding purpose-fit computation schedules for networks. We define a novel family of “clockwork” convnets driven by fixed or adaptive clock signals that schedule the processing of different layers at different update rates according to their semantic stability. We design a pipeline schedule to reduce latency for real-time recognition and a fixed-rate schedule to reduce overall computation. Finally, we extend clockwork scheduling to adaptive video processing by incorporating data-driven clocks that can be tuned on unlabeled video. The accuracy and efficiency of clockwork convnets are evaluated on the Youtube-Objects, NYUD, and Cityscapes video datasets. |
| Tasks | Semantic Segmentation, Video Recognition, Video Semantic Segmentation |
| Published | 2016-08-11 |
| URL | http://arxiv.org/abs/1608.03609v1 |
| http://arxiv.org/pdf/1608.03609v1.pdf | |
| PWC | https://paperswithcode.com/paper/clockwork-convnets-for-video-semantic |
| Repo | https://github.com/shelhamer/clockwork-fcn |
| Framework | caffe2 |
RL$^2$: Fast Reinforcement Learning via Slow Reinforcement Learning
| Title | RL$^2$: Fast Reinforcement Learning via Slow Reinforcement Learning |
| Authors | Yan Duan, John Schulman, Xi Chen, Peter L. Bartlett, Ilya Sutskever, Pieter Abbeel |
| Abstract | Deep reinforcement learning (deep RL) has been successful in learning sophisticated behaviors automatically; however, the learning process requires a huge number of trials. In contrast, animals can learn new tasks in just a few trials, benefiting from their prior knowledge about the world. This paper seeks to bridge this gap. Rather than designing a “fast” reinforcement learning algorithm, we propose to represent it as a recurrent neural network (RNN) and learn it from data. In our proposed method, RL$^2$, the algorithm is encoded in the weights of the RNN, which are learned slowly through a general-purpose (“slow”) RL algorithm. The RNN receives all information a typical RL algorithm would receive, including observations, actions, rewards, and termination flags; and it retains its state across episodes in a given Markov Decision Process (MDP). The activations of the RNN store the state of the “fast” RL algorithm on the current (previously unseen) MDP. We evaluate RL$^2$ experimentally on both small-scale and large-scale problems. On the small-scale side, we train it to solve randomly generated multi-arm bandit problems and finite MDPs. After RL$^2$ is trained, its performance on new MDPs is close to human-designed algorithms with optimality guarantees. On the large-scale side, we test RL$^2$ on a vision-based navigation task and show that it scales up to high-dimensional problems. |
| Tasks | |
| Published | 2016-11-09 |
| URL | http://arxiv.org/abs/1611.02779v2 |
| http://arxiv.org/pdf/1611.02779v2.pdf | |
| PWC | https://paperswithcode.com/paper/rl2-fast-reinforcement-learning-via-slow |
| Repo | https://github.com/dragen1860/MAML-Pytorch-RL |
| Framework | pytorch |
Instance Normalization: The Missing Ingredient for Fast Stylization
| Title | Instance Normalization: The Missing Ingredient for Fast Stylization |
| Authors | Dmitry Ulyanov, Andrea Vedaldi, Victor Lempitsky |
| Abstract | It this paper we revisit the fast stylization method introduced in Ulyanov et. al. (2016). We show how a small change in the stylization architecture results in a significant qualitative improvement in the generated images. The change is limited to swapping batch normalization with instance normalization, and to apply the latter both at training and testing times. The resulting method can be used to train high-performance architectures for real-time image generation. The code will is made available on github at https://github.com/DmitryUlyanov/texture_nets. Full paper can be found at arXiv:1701.02096. |
| Tasks | Image Generation, Image Stylization, Style Transfer |
| Published | 2016-07-27 |
| URL | http://arxiv.org/abs/1607.08022v3 |
| http://arxiv.org/pdf/1607.08022v3.pdf | |
| PWC | https://paperswithcode.com/paper/instance-normalization-the-missing-ingredient |
| Repo | https://github.com/tbullmann/imagetranslation-tensorflow |
| Framework | tf |
A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks
| Title | A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks |
| Authors | Dan Hendrycks, Kevin Gimpel |
| Abstract | We consider the two related problems of detecting if an example is misclassified or out-of-distribution. We present a simple baseline that utilizes probabilities from softmax distributions. Correctly classified examples tend to have greater maximum softmax probabilities than erroneously classified and out-of-distribution examples, allowing for their detection. We assess performance by defining several tasks in computer vision, natural language processing, and automatic speech recognition, showing the effectiveness of this baseline across all. We then show the baseline can sometimes be surpassed, demonstrating the room for future research on these underexplored detection tasks. |
| Tasks | Speech Recognition |
| Published | 2016-10-07 |
| URL | http://arxiv.org/abs/1610.02136v3 |
| http://arxiv.org/pdf/1610.02136v3.pdf | |
| PWC | https://paperswithcode.com/paper/a-baseline-for-detecting-misclassified-and |
| Repo | https://github.com/dabsdamoon/MNIST-Auxiliary-Decoder |
| Framework | none |
IRLS and Slime Mold: Equivalence and Convergence
| Title | IRLS and Slime Mold: Equivalence and Convergence |
| Authors | Damian Straszak, Nisheeth K. Vishnoi |
| Abstract | In this paper we present a connection between two dynamical systems arising in entirely different contexts: one in signal processing and the other in biology. The first is the famous Iteratively Reweighted Least Squares (IRLS) algorithm used in compressed sensing and sparse recovery while the second is the dynamics of a slime mold (Physarum polycephalum). Both of these dynamics are geared towards finding a minimum l1-norm solution in an affine subspace. Despite its simplicity the convergence of the IRLS method has been shown only for a certain regularization of it and remains an important open problem. Our first result shows that the two dynamics are projections of the same dynamical system in higher dimensions. As a consequence, and building on the recent work on Physarum dynamics, we are able to prove convergence and obtain complexity bounds for a damped version of the IRLS algorithm. |
| Tasks | |
| Published | 2016-01-12 |
| URL | http://arxiv.org/abs/1601.02712v1 |
| http://arxiv.org/pdf/1601.02712v1.pdf | |
| PWC | https://paperswithcode.com/paper/irls-and-slime-mold-equivalence-and |
| Repo | https://github.com/DamianStraszak/IRLS-and-Physarum-Dynamics |
| Framework | none |
Real-time Action Recognition with Enhanced Motion Vector CNNs
| Title | Real-time Action Recognition with Enhanced Motion Vector CNNs |
| Authors | Bowen Zhang, Limin Wang, Zhe Wang, Yu Qiao, Hanli Wang |
| Abstract | The deep two-stream architecture exhibited excellent performance on video based action recognition. The most computationally expensive step in this approach comes from the calculation of optical flow which prevents it to be real-time. This paper accelerates this architecture by replacing optical flow with motion vector which can be obtained directly from compressed videos without extra calculation. However, motion vector lacks fine structures, and contains noisy and inaccurate motion patterns, leading to the evident degradation of recognition performance. Our key insight for relieving this problem is that optical flow and motion vector are inherent correlated. Transferring the knowledge learned with optical flow CNN to motion vector CNN can significantly boost the performance of the latter. Specifically, we introduce three strategies for this, initialization transfer, supervision transfer and their combination. Experimental results show that our method achieves comparable recognition performance to the state-of-the-art, while our method can process 390.7 frames per second, which is 27 times faster than the original two-stream method. |
| Tasks | Optical Flow Estimation, Temporal Action Localization |
| Published | 2016-04-26 |
| URL | http://arxiv.org/abs/1604.07669v1 |
| http://arxiv.org/pdf/1604.07669v1.pdf | |
| PWC | https://paperswithcode.com/paper/real-time-action-recognition-with-enhanced |
| Repo | https://github.com/yjxiong/caffe |
| Framework | none |
High-Resolution Image Inpainting using Multi-Scale Neural Patch Synthesis
| Title | High-Resolution Image Inpainting using Multi-Scale Neural Patch Synthesis |
| Authors | Chao Yang, Xin Lu, Zhe Lin, Eli Shechtman, Oliver Wang, Hao Li |
| Abstract | Recent advances in deep learning have shown exciting promise in filling large holes in natural images with semantically plausible and context aware details, impacting fundamental image manipulation tasks such as object removal. While these learning-based methods are significantly more effective in capturing high-level features than prior techniques, they can only handle very low-resolution inputs due to memory limitations and difficulty in training. Even for slightly larger images, the inpainted regions would appear blurry and unpleasant boundaries become visible. We propose a multi-scale neural patch synthesis approach based on joint optimization of image content and texture constraints, which not only preserves contextual structures but also produces high-frequency details by matching and adapting patches with the most similar mid-layer feature correlations of a deep classification network. We evaluate our method on the ImageNet and Paris Streetview datasets and achieved state-of-the-art inpainting accuracy. We show our approach produces sharper and more coherent results than prior methods, especially for high-resolution images. |
| Tasks | Image Inpainting |
| Published | 2016-11-30 |
| URL | http://arxiv.org/abs/1611.09969v2 |
| http://arxiv.org/pdf/1611.09969v2.pdf | |
| PWC | https://paperswithcode.com/paper/high-resolution-image-inpainting-using-multi |
| Repo | https://github.com/leehomyc/Faster-High-Res-Neural-Inpainting |
| Framework | torch |
Finding Better Active Learners for Faster Literature Reviews
| Title | Finding Better Active Learners for Faster Literature Reviews |
| Authors | Zhe Yu, Nicholas A. Kraft, Tim Menzies |
| Abstract | Literature reviews can be time-consuming and tedious to complete. By cataloging and refactoring three state-of-the-art active learning techniques from evidence-based medicine and legal electronic discovery, this paper finds and implements FASTREAD, a faster technique for studying a large corpus of documents. This paper assesses FASTREAD using datasets generated from existing SE literature reviews (Hall, Wahono, Radjenovi'c, Kitchenham et al.). Compared to manual methods, FASTREAD lets researchers find 95% relevant studies after reviewing an order of magnitude fewer papers. Compared to other state-of-the-art automatic methods, FASTREAD reviews 20-50% fewer studies while finding same number of relevant primary studies in a systematic literature review. |
| Tasks | Active Learning |
| Published | 2016-12-10 |
| URL | http://arxiv.org/abs/1612.03224v5 |
| http://arxiv.org/pdf/1612.03224v5.pdf | |
| PWC | https://paperswithcode.com/paper/finding-better-active-learners-for-faster |
| Repo | https://github.com/fastread/src |
| Framework | none |
MPI-FAUN: An MPI-Based Framework for Alternating-Updating Nonnegative Matrix Factorization
| Title | MPI-FAUN: An MPI-Based Framework for Alternating-Updating Nonnegative Matrix Factorization |
| Authors | Ramakrishnan Kannan, Grey Ballard, Haesun Park |
| Abstract | Non-negative matrix factorization (NMF) is the problem of determining two non-negative low rank factors $W$ and $H$, for the given input matrix $A$, such that $A \approx W H$. NMF is a useful tool for many applications in different domains such as topic modeling in text mining, background separation in video analysis, and community detection in social networks. Despite its popularity in the data mining community, there is a lack of efficient parallel algorithms to solve the problem for big data sets. The main contribution of this work is a new, high-performance parallel computational framework for a broad class of NMF algorithms that iteratively solves alternating non-negative least squares (NLS) subproblems for $W$ and $H$. It maintains the data and factor matrices in memory (distributed across processors), uses MPI for interprocessor communication, and, in the dense case, provably minimizes communication costs (under mild assumptions). The framework is flexible and able to leverage a variety of NMF and NLS algorithms, including Multiplicative Update, Hierarchical Alternating Least Squares, and Block Principal Pivoting. Our implementation allows us to benchmark and compare different algorithms on massive dense and sparse data matrices of size that spans for few hundreds of millions to billions. We demonstrate the scalability of our algorithm and compare it with baseline implementations, showing significant performance improvements. The code and the datasets used for conducting the experiments are available online. |
| Tasks | Community Detection |
| Published | 2016-09-28 |
| URL | http://arxiv.org/abs/1609.09154v1 |
| http://arxiv.org/pdf/1609.09154v1.pdf | |
| PWC | https://paperswithcode.com/paper/mpi-faun-an-mpi-based-framework-for |
| Repo | https://github.com/ramkikannan/nmflibrary |
| Framework | none |
An Actor-Critic Algorithm for Sequence Prediction
| Title | An Actor-Critic Algorithm for Sequence Prediction |
| Authors | Dzmitry Bahdanau, Philemon Brakel, Kelvin Xu, Anirudh Goyal, Ryan Lowe, Joelle Pineau, Aaron Courville, Yoshua Bengio |
| Abstract | We present an approach to training neural networks to generate sequences using actor-critic methods from reinforcement learning (RL). Current log-likelihood training methods are limited by the discrepancy between their training and testing modes, as models must generate tokens conditioned on their previous guesses rather than the ground-truth tokens. We address this problem by introducing a \textit{critic} network that is trained to predict the value of an output token, given the policy of an \textit{actor} network. This results in a training procedure that is much closer to the test phase, and allows us to directly optimize for a task-specific score such as BLEU. Crucially, since we leverage these techniques in the supervised learning setting rather than the traditional RL setting, we condition the critic network on the ground-truth output. We show that our method leads to improved performance on both a synthetic task, and for German-English machine translation. Our analysis paves the way for such methods to be applied in natural language generation tasks, such as machine translation, caption generation, and dialogue modelling. |
| Tasks | Machine Translation, Spelling Correction, Text Generation |
| Published | 2016-07-24 |
| URL | http://arxiv.org/abs/1607.07086v3 |
| http://arxiv.org/pdf/1607.07086v3.pdf | |
| PWC | https://paperswithcode.com/paper/an-actor-critic-algorithm-for-sequence |
| Repo | https://github.com/juliakreutzer/joeynmt |
| Framework | pytorch |
Quasi-Recurrent Neural Networks
| Title | Quasi-Recurrent Neural Networks |
| Authors | James Bradbury, Stephen Merity, Caiming Xiong, Richard Socher |
| Abstract | Recurrent neural networks are a powerful tool for modeling sequential data, but the dependence of each timestep’s computation on the previous timestep’s output limits parallelism and makes RNNs unwieldy for very long sequences. We introduce quasi-recurrent neural networks (QRNNs), an approach to neural sequence modeling that alternates convolutional layers, which apply in parallel across timesteps, and a minimalist recurrent pooling function that applies in parallel across channels. Despite lacking trainable recurrent layers, stacked QRNNs have better predictive accuracy than stacked LSTMs of the same hidden size. Due to their increased parallelism, they are up to 16 times faster at train and test time. Experiments on language modeling, sentiment classification, and character-level neural machine translation demonstrate these advantages and underline the viability of QRNNs as a basic building block for a variety of sequence tasks. |
| Tasks | Language Modelling, Machine Translation, Sentiment Analysis |
| Published | 2016-11-05 |
| URL | http://arxiv.org/abs/1611.01576v2 |
| http://arxiv.org/pdf/1611.01576v2.pdf | |
| PWC | https://paperswithcode.com/paper/quasi-recurrent-neural-networks |
| Repo | https://github.com/salesforce/pytorch-qrnn |
| Framework | pytorch |
Authorship clustering using multi-headed recurrent neural networks
| Title | Authorship clustering using multi-headed recurrent neural networks |
| Authors | Douglas Bagnall |
| Abstract | A recurrent neural network that has been trained to separately model the language of several documents by unknown authors is used to measure similarity between the documents. It is able to find clues of common authorship even when the documents are very short and about disparate topics. While it is easy to make statistically significant predictions regarding authorship, it is difficult to group documents into definite clusters with high accuracy. |
| Tasks | |
| Published | 2016-08-16 |
| URL | http://arxiv.org/abs/1608.04485v1 |
| http://arxiv.org/pdf/1608.04485v1.pdf | |
| PWC | https://paperswithcode.com/paper/authorship-clustering-using-multi-headed |
| Repo | https://github.com/douglasbagnall/bog |
| Framework | none |
Memory-Efficient Backpropagation Through Time
| Title | Memory-Efficient Backpropagation Through Time |
| Authors | Audrūnas Gruslys, Remi Munos, Ivo Danihelka, Marc Lanctot, Alex Graves |
| Abstract | We propose a novel approach to reduce memory consumption of the backpropagation through time (BPTT) algorithm when training recurrent neural networks (RNNs). Our approach uses dynamic programming to balance a trade-off between caching of intermediate results and recomputation. The algorithm is capable of tightly fitting within almost any user-set memory budget while finding an optimal execution policy minimizing the computational cost. Computational devices have limited memory capacity and maximizing a computational performance given a fixed memory budget is a practical use-case. We provide asymptotic computational upper bounds for various regimes. The algorithm is particularly effective for long sequences. For sequences of length 1000, our algorithm saves 95% of memory usage while using only one third more time per iteration than the standard BPTT. |
| Tasks | |
| Published | 2016-06-10 |
| URL | http://arxiv.org/abs/1606.03401v1 |
| http://arxiv.org/pdf/1606.03401v1.pdf | |
| PWC | https://paperswithcode.com/paper/memory-efficient-backpropagation-through-time |
| Repo | https://github.com/cybertronai/gradient-checkpointing |
| Framework | tf |
Logic Tensor Networks: Deep Learning and Logical Reasoning from Data and Knowledge
| Title | Logic Tensor Networks: Deep Learning and Logical Reasoning from Data and Knowledge |
| Authors | Luciano Serafini, Artur d’Avila Garcez |
| Abstract | We propose Logic Tensor Networks: a uniform framework for integrating automatic learning and reasoning. A logic formalism called Real Logic is defined on a first-order language whereby formulas have truth-value in the interval [0,1] and semantics defined concretely on the domain of real numbers. Logical constants are interpreted as feature vectors of real numbers. Real Logic promotes a well-founded integration of deductive reasoning on a knowledge-base and efficient data-driven relational machine learning. We show how Real Logic can be implemented in deep Tensor Neural Networks with the use of Google’s tensorflow primitives. The paper concludes with experiments applying Logic Tensor Networks on a simple but representative example of knowledge completion. |
| Tasks | Tensor Networks |
| Published | 2016-06-14 |
| URL | http://arxiv.org/abs/1606.04422v2 |
| http://arxiv.org/pdf/1606.04422v2.pdf | |
| PWC | https://paperswithcode.com/paper/logic-tensor-networks-deep-learning-and |
| Repo | https://github.com/thadumi/logictensornetworks-keras |
| Framework | tf |
