Paper Group ANR 1127
Efficient Decremental Learning Algorithms for Broad Learning System. UWB-GCN: Accelerating Graph Convolutional Networks through Runtime Workload Rebalancing. Learning Temporal Action Proposals With Fewer Labels. Policies for constraining the behaviour of coalitions of agents in the context of algebraic information theory. Foothill: A Quasiconvex Re …
Efficient Decremental Learning Algorithms for Broad Learning System
| Title | Efficient Decremental Learning Algorithms for Broad Learning System |
| Authors | Hufei Zhu |
| Abstract | The decremented learning algorithms are required in machine learning, to prune redundant nodes and remove obsolete inline training samples. In this paper, an efficient decremented learning algorithm to prune redundant nodes is deduced from the incremental learning algorithm 1 proposed in [9] for added nodes, and two decremented learning algorithms to remove training samples are deduced from the two incremental learning algorithms proposed in [10] for added inputs. The proposed decremented learning algorithm for reduced nodes utilizes the inverse Cholesterol factor of the Herminia matrix in the ridge inverse, to update the output weights recursively, as the incremental learning algorithm 1 for added nodes in [9], while that inverse Cholesterol factor is updated with an unitary transformation. The proposed decremented learning algorithm 1 for reduced inputs updates the output weights recursively with the inverse of the Herminia matrix in the ridge inverse, and updates that inverse recursively, as the incremental learning algorithm 1 for added inputs in [10]. |
| Tasks | |
| Published | 2019-12-31 |
| URL | https://arxiv.org/abs/1912.13169v1 |
| https://arxiv.org/pdf/1912.13169v1.pdf | |
| PWC | https://paperswithcode.com/paper/efficient-decremental-learning-algorithms-for |
| Repo | |
| Framework | |
UWB-GCN: Accelerating Graph Convolutional Networks through Runtime Workload Rebalancing
| Title | UWB-GCN: Accelerating Graph Convolutional Networks through Runtime Workload Rebalancing |
| Authors | Tong Geng, Ang Li, Tianqi Wang, Chunshu Wu, Yanfei Li, Antonino Tumeo, Shuai Che, Steve Reinhardt, Martin Herbordt |
| Abstract | Deep learning systems have been applied mostly to Euclidean data such as images, video, and audio. In many applications, however, information and their relationships are better expressed with graphs. Graph convolutional networks (GCNs) appear to be a promising approach to efficiently learn from graph data structures, having shown advantages in many critical applications such as power system analysis, chemical reactivity prediction, material property prediction, E-commerce, cybersecurity, etc. As with other deep learning modalities, hardware acceleration is critical. The challenge is that real-world graphs are often extremely large and unbalanced; this poses both significant performance demands and design challenges. We propose an architecture that accelerates GCN inference, the Ultra Workload Balanced GCN (UWB-GCN). To address the major performance bottleneck of workload imbalance we propose two techniques, dynamic local sharing and dynamic remote switching. Both rely on hardware flexibility to autotune the system; this is effected with the negligible area and delay overhead. In particular, UWB-GCN profiles the sparse graph pattern while continuously adjusting the workload distribution strategy among a large number of processing elements (PEs). Once the system converges to an ideal configuration, this configuration is used for the remaining iterations. To the best of our knowledge, UWB-GCN is the first accelerator design targeting GCNs and the first that autotunes workload balance in the accelerator in hardware rather than software. These methods result in a near-ideal workload balance in processing sparse matrices. Experimental results show that UWB-GCN can perform inference of the Nell graph (66K vertices, 266K edges) in 8.1 ms; this corresponds to speedups of 199x, 16x, and 7.5x as compared with, respectively, CPU, GPU, and a baseline design with no workload rebalancing. |
| Tasks | |
| Published | 2019-08-23 |
| URL | https://arxiv.org/abs/1908.10834v4 |
| https://arxiv.org/pdf/1908.10834v4.pdf | |
| PWC | https://paperswithcode.com/paper/uwb-gcn-hardware-acceleration-of-graph |
| Repo | |
| Framework | |
Learning Temporal Action Proposals With Fewer Labels
| Title | Learning Temporal Action Proposals With Fewer Labels |
| Authors | Jingwei Ji, Kaidi Cao, Juan Carlos Niebles |
| Abstract | Temporal action proposals are a common module in action detection pipelines today. Most current methods for training action proposal modules rely on fully supervised approaches that require large amounts of annotated temporal action intervals in long video sequences. The large cost and effort in annotation that this entails motivate us to study the problem of training proposal modules with less supervision. In this work, we propose a semi-supervised learning algorithm specifically designed for training temporal action proposal networks. When only a small number of labels are available, our semi-supervised method generates significantly better proposals than the fully-supervised counterpart and other strong semi-supervised baselines. We validate our method on two challenging action detection video datasets, ActivityNet v1.3 and THUMOS14. We show that our semi-supervised approach consistently matches or outperforms the fully supervised state-of-the-art approaches. |
| Tasks | Action Detection |
| Published | 2019-10-03 |
| URL | https://arxiv.org/abs/1910.01286v1 |
| https://arxiv.org/pdf/1910.01286v1.pdf | |
| PWC | https://paperswithcode.com/paper/learning-temporal-action-proposals-with-fewer |
| Repo | |
| Framework | |
Policies for constraining the behaviour of coalitions of agents in the context of algebraic information theory
| Title | Policies for constraining the behaviour of coalitions of agents in the context of algebraic information theory |
| Authors | Christopher Goddard |
| Abstract | This article takes an oblique sidestep from two previous papers, wherein an approach to reformulation of game theory in terms of information theory, topology, as well as a few other notions was indicated. In this document a description is provided as to how one might determine an approach for an agent to choose a policy concerning which actions to take in a game that constrains behaviour of subsidiary agents. It is then demonstrated how these results in algebraic information theory, together with previous investigations in geometric and topological information theory, can be unified into a single cohesive framework. |
| Tasks | |
| Published | 2019-11-28 |
| URL | https://arxiv.org/abs/1912.00803v1 |
| https://arxiv.org/pdf/1912.00803v1.pdf | |
| PWC | https://paperswithcode.com/paper/policies-for-constraining-the-behaviour-of |
| Repo | |
| Framework | |
Foothill: A Quasiconvex Regularization for Edge Computing of Deep Neural Networks
| Title | Foothill: A Quasiconvex Regularization for Edge Computing of Deep Neural Networks |
| Authors | Mouloud Belbahri, Eyyüb Sari, Sajad Darabi, Vahid Partovi Nia |
| Abstract | Deep neural networks (DNNs) have demonstrated success for many supervised learning tasks, ranging from voice recognition, object detection, to image classification. However, their increasing complexity might yield poor generalization error that make them hard to be deployed on edge devices. Quantization is an effective approach to compress DNNs in order to meet these constraints. Using a quasiconvex base function in order to construct a binary quantizer helps training binary neural networks (BNNs) and adding noise to the input data or using a concrete regularization function helps to improve generalization error. Here we introduce foothill function, an infinitely differentiable quasiconvex function. This regularizer is flexible enough to deform towards $L_1$ and $L_2$ penalties. Foothill can be used as a binary quantizer, as a regularizer, or as a loss. In particular, we show this regularizer reduces the accuracy gap between BNNs and their full-precision counterpart for image classification on ImageNet. |
| Tasks | Image Classification, Object Detection, Quantization |
| Published | 2019-01-18 |
| URL | https://arxiv.org/abs/1901.06414v2 |
| https://arxiv.org/pdf/1901.06414v2.pdf | |
| PWC | https://paperswithcode.com/paper/foothill-a-quasiconvex-regularization |
| Repo | |
| Framework | |
Evaluation Metrics for Unsupervised Learning Algorithms
| Title | Evaluation Metrics for Unsupervised Learning Algorithms |
| Authors | Julio-Omar Palacio-Niño, Fernando Berzal |
| Abstract | Determining the quality of the results obtained by clustering techniques is a key issue in unsupervised machine learning. Many authors have discussed the desirable features of good clustering algorithms. However, Jon Kleinberg established an impossibility theorem for clustering. As a consequence, a wealth of studies have proposed techniques to evaluate the quality of clustering results depending on the characteristics of the clustering problem and the algorithmic technique employed to cluster data. |
| Tasks | |
| Published | 2019-05-14 |
| URL | https://arxiv.org/abs/1905.05667v2 |
| https://arxiv.org/pdf/1905.05667v2.pdf | |
| PWC | https://paperswithcode.com/paper/evaluation-metrics-for-unsupervised-learning |
| Repo | |
| Framework | |
Delving into VoxCeleb: environment invariant speaker recognition
| Title | Delving into VoxCeleb: environment invariant speaker recognition |
| Authors | Joon Son Chung, Jaesung Huh, Seongkyu Mun |
| Abstract | Research in speaker recognition has recently seen significant progress due to the application of neural network models and the availability of new large-scale datasets. There has been a plethora of work in search for more powerful architectures or loss functions suitable for the task, but these works do not consider what information is learnt by the models, apart from being able to predict the given labels. In this work, we introduce an environment adversarial training framework in which the network can effectively learn speaker-discriminative and environment-invariant embeddings without explicit domain shift during training. We achieve this by utilising the previously unused `video’ information in the VoxCeleb dataset. The environment adversarial training allows the network to generalise better to unseen conditions. The method is evaluated on both speaker identification and verification tasks using the VoxCeleb dataset, on which we demonstrate significant performance improvements over baselines. | |
| Tasks | Speaker Identification, Speaker Recognition |
| Published | 2019-10-24 |
| URL | https://arxiv.org/abs/1910.11238v2 |
| https://arxiv.org/pdf/1910.11238v2.pdf | |
| PWC | https://paperswithcode.com/paper/delving-into-voxceleb-environment-invariant |
| Repo | |
| Framework | |
4D MRI: Robust sorting of free breathing MRI slices for use in interventional settings
| Title | 4D MRI: Robust sorting of free breathing MRI slices for use in interventional settings |
| Authors | Gino Gulamhussene, Fabian Joeres, Marko Rak, Maciej Pech, Christian Hansen |
| Abstract | Purpose: We aim to develop a robust 4D MRI method for large FOVs enabling the extraction of irregular respiratory motion that is readily usable with all MRI machines and thus applicable to support a wide range of interventional settings. Method: We propose a 4D MRI reconstruction method to capture an arbitrary number of breathing states. It uses template updates in navigator slices and search regions for fast and robust vessel cross-section tracking. It captures FOVs of 255 mm x 320 mm x 228 mm at a spatial resolution of 1.82 mm x 1.82 mm x 4mm and temporal resolution of 200ms. To validate the method, a total of 38 4D MRIs of 13 healthy subjects were reconstructed. A quantitative evaluation of the reconstruction rate and speed of both the new and baseline method was performed. Additionally, a study with ten radiologists was conducted to assess the subjective reconstruction quality of both methods. Results: Our results indicate improved mean reconstruction rates compared to the baseline method (79.4% vs. 45.5%) and improved mean reconstruction times (24s vs. 73s) per subject. Interventional radiologists perceive the reconstruction quality of our method as higher compared to the baseline (262.5 points vs. 217.5 points, p=0.02). Conclusions: Template updates are an effective and efficient way to increase 4D MRI reconstruction rates and to achieve better reconstruction quality. Search regions reduce reconstruction time. These improvements increase the applicability of 4D MRI as base for seamless support of interventional image guidance in percutaneous interventions. |
| Tasks | |
| Published | 2019-10-04 |
| URL | https://arxiv.org/abs/1910.01902v1 |
| https://arxiv.org/pdf/1910.01902v1.pdf | |
| PWC | https://paperswithcode.com/paper/4d-mri-robust-sorting-of-free-breathing-mri |
| Repo | |
| Framework | |
On Policy Learning Robust to Irreversible Events: An Application to Robotic In-Hand Manipulation
| Title | On Policy Learning Robust to Irreversible Events: An Application to Robotic In-Hand Manipulation |
| Authors | Pietro Falco, Abdallah Attawia, Matteo Saveriano, Dongheui Lee |
| Abstract | In this letter, we present an approach for learning in-hand manipulation skills with a low-cost, underactuated prosthetic hand in the presence of irreversible events. Our approach combines reinforcement learning based on visual perception with low-level reactive control based on tactile perception, which aims to avoid slipping. The objective of the reinforcement learning level consists not only in fulfilling the in-hand manipulation goal, but also in minimizing the intervention of the tactile reactive control. This way, the occurrence of object slipping during the learning procedure, which we consider an irreversible event, is significantly reduced. When an irreversible event occurs, the learning process is considered failed. We show the performance in two tasks, which consist in reorienting a cup and a bottle only using the fingers. The experimental results show that the proposed architecture allows reaching the goal in the Cartesian space and reduces significantly the occurrence of object slipping during the learning procedure. Moreover, without the proposed synergy between reactive control and reinforcement learning it was not possible to avoid irreversible events and, therefore, to learn the task. |
| Tasks | |
| Published | 2019-11-20 |
| URL | https://arxiv.org/abs/1911.08927v1 |
| https://arxiv.org/pdf/1911.08927v1.pdf | |
| PWC | https://paperswithcode.com/paper/on-policy-learning-robust-to-irreversible |
| Repo | |
| Framework | |
Primary User Emulation Attacks: A Detection Technique Based on Kalman Filter
| Title | Primary User Emulation Attacks: A Detection Technique Based on Kalman Filter |
| Authors | Zakaria El Mrabet, Youness Arjoune, Hassan El Ghazi, Badr Abou Al Majd, Naima Kaabouch |
| Abstract | Cognitive radio technology addresses the problem of spectrum scarcity by allowing secondary users to use the vacant spectrum bands without causing interference to the primary users. However, several attacks could disturb the normal functioning of the cognitive radio network. Primary user emulation attacks are one of the most severe attacks in which a malicious user emulates the primary user signal characteristics to either prevent other legitimate secondary users from accessing the idle channels or causing harmful interference to the primary users. There are several proposed approaches to detect the primary user emulation attackers. However, most of these techniques assume that the primary user location is fixed, which does not make them valid when the primary user is mobile. In this paper, we propose a new approach based on the Kalman filter framework for detecting the primary user emulation attacks with a non-stationary primary user. Several experiments have been conducted and the advantages of the proposed approach are demonstrated through the simulation results. |
| Tasks | |
| Published | 2019-03-08 |
| URL | http://arxiv.org/abs/1903.03684v1 |
| http://arxiv.org/pdf/1903.03684v1.pdf | |
| PWC | https://paperswithcode.com/paper/primary-user-emulation-attacks-a-detection |
| Repo | |
| Framework | |
Worst-Case Regret Bounds for Exploration via Randomized Value Functions
| Title | Worst-Case Regret Bounds for Exploration via Randomized Value Functions |
| Authors | Daniel Russo |
| Abstract | This paper studies a recent proposal to use randomized value functions to drive exploration in reinforcement learning. These randomized value functions are generated by injecting random noise into the training data, making the approach compatible with many popular methods for estimating parameterized value functions. By providing a worst-case regret bound for tabular finite-horizon Markov decision processes, we show that planning with respect to these randomized value functions can induce provably efficient exploration. |
| Tasks | Efficient Exploration |
| Published | 2019-06-07 |
| URL | https://arxiv.org/abs/1906.02870v3 |
| https://arxiv.org/pdf/1906.02870v3.pdf | |
| PWC | https://paperswithcode.com/paper/worst-case-regret-bounds-for-exploration-via |
| Repo | |
| Framework | |
A Simple yet Effective Way for Improving the Performance of GANs
| Title | A Simple yet Effective Way for Improving the Performance of GANs |
| Authors | Yong-Goo Shin, Yoon-Jae Yeo, Min-Cheol Sagong, Cheol-Hwan Yoo, Sung-Jea Ko |
| Abstract | This paper presents a simple but effective way that improves the performance of generative adversarial networks (GANs) without imposing the training overhead or modifying the network architectures of existing methods. The proposed method employs a novel cascading rejection (CR) module for discriminator, which extracts multiple non-overlapped features in an iterative manner. The CR module supports the discriminator to effectively distinguish between real and generated images, which results in a strong penalization to the generator. In order to deceive the robust discriminator containing the CR module, the generator produces the images that are more similar to the real images. Since the proposed CR module requires only a few simple vector operations, it can be readily applied to existing frameworks with marginal training overheads. Quantitative evaluations on various datasets including CIFAR-10, Celeb-HQ, LSUN, and tiny-ImageNet confirm that the proposed method significantly improves the performance of GANs and conditional GANs in terms of Frechet inception distance (FID) indicating the diversity and visual appearance of the generated images. |
| Tasks | |
| Published | 2019-11-19 |
| URL | https://arxiv.org/abs/1911.10979v2 |
| https://arxiv.org/pdf/1911.10979v2.pdf | |
| PWC | https://paperswithcode.com/paper/a-simple-yet-effective-way-for-improving-the |
| Repo | |
| Framework | |
Optimizing Routerless Network-on-Chip Designs: An Innovative Learning-Based Framework
| Title | Optimizing Routerless Network-on-Chip Designs: An Innovative Learning-Based Framework |
| Authors | Ting-Ru Lin, Drew Penney, Massoud Pedram, Lizhong Chen |
| Abstract | Machine learning applied to architecture design presents a promising opportunity with broad applications. Recent deep reinforcement learning (DRL) techniques, in particular, enable efficient exploration in vast design spaces where conventional design strategies may be inadequate. This paper proposes a novel deep reinforcement framework, taking routerless networks-on-chip (NoC) as an evaluation case study. The new framework successfully resolves problems with prior design approaches being either unreliable due to random searches or inflexible due to severe design space restrictions. The framework learns (near-)optimal loop placement for routerless NoCs with various design constraints. A deep neural network is developed using parallel threads that efficiently explore the immense routerless NoC design space with a Monte Carlo search tree. Experimental results show that, compared with conventional mesh, the proposed deep reinforcement learning (DRL) routerless design achieves a 3.25x increase in throughput, 1.6x reduction in packet latency, and 5x reduction in power. Compared with the state-of-the-art routerless NoC, DRL achieves a 1.47x increase in throughput, 1.18x reduction in packet latency, and 1.14x reduction in average hop count albeit with slightly more power overhead. |
| Tasks | Efficient Exploration |
| Published | 2019-05-11 |
| URL | https://arxiv.org/abs/1905.04423v1 |
| https://arxiv.org/pdf/1905.04423v1.pdf | |
| PWC | https://paperswithcode.com/paper/optimizing-routerless-network-on-chip-designs |
| Repo | |
| Framework | |
Latent feature disentanglement for 3D meshes
| Title | Latent feature disentanglement for 3D meshes |
| Authors | Jake Levinson, Avneesh Sud, Ameesh Makadia |
| Abstract | Generative modeling of 3D shapes has become an important problem due to its relevance to many applications across Computer Vision, Graphics, and VR. In this paper we build upon recently introduced 3D mesh-convolutional Variational AutoEncoders which have shown great promise for learning rich representations of deformable 3D shapes. We introduce a supervised generative 3D mesh model that disentangles the latent shape representation into independent generative factors. Our extensive experimental analysis shows that learning an explicitly disentangled representation can both improve random shape generation as well as successfully address downstream tasks such as pose and shape transfer, shape-invariant temporal synchronization, and pose-invariant shape matching. |
| Tasks | |
| Published | 2019-06-07 |
| URL | https://arxiv.org/abs/1906.03281v1 |
| https://arxiv.org/pdf/1906.03281v1.pdf | |
| PWC | https://paperswithcode.com/paper/latent-feature-disentanglement-for-3d-meshes |
| Repo | |
| Framework | |
Evaluation of head segmentation quality for treatment planning of tumor treating fields in brain tumors
| Title | Evaluation of head segmentation quality for treatment planning of tumor treating fields in brain tumors |
| Authors | Reuben R Shamir, Zeev Bomzon |
| Abstract | Tumor treating fields (TTFields) is an FDA approved therapy for the treatment of Gliobastoma Multiform (GBM) and currently being investigated for additional tumor types. TTFields are delivered to the tumor through the placement of transducer arrays (TAs) placed on the patient scalp. The positions of the TAs are associated with treatment outcomes via simulations of the electric fields. Therefore, we are currently developing a method for recommending optimal placement of TAs. A key step to achieve this goal is to correctly segment the head into tissues of similar electrical properties. Visual inspection of segmentation quality is invaluable but time-consuming. Automatic quality assessment can assist in automatic refinement of the segmentation parameters, suggest flaw points to the user and indicate if the segmented method is of sufficient accuracy for TTFields simulation. As a first step in this direction, we identified a set of features that are relevant to atlas-based segmentation and show that these are significantly correlated (p < 0.05) with a similarity measure between validated and automatically computed segmentations. Furthermore, we incorporated these features in a decision tree regressor to predict the similarity of the validated and computed segmentations of 20 TTFields patients using a leave-one-out approach. The predicted similarity measures were highly correlated with the actual ones (average abs. difference 3% (SD = 3%); r = 0.92, p < 0.001). We conclude that quality estimation of segmentations is feasible by incorporating machine learning and segmentation-relevant features. |
| Tasks | |
| Published | 2019-06-26 |
| URL | https://arxiv.org/abs/1906.11014v1 |
| https://arxiv.org/pdf/1906.11014v1.pdf | |
| PWC | https://paperswithcode.com/paper/evaluation-of-head-segmentation-quality-for |
| Repo | |
| Framework | |
