Filters








241 Hits in 5.1 sec

LCS: Learning Compressible Subspaces for Adaptive Network Compression at Inference Time [article]

Elvis Nunez, Maxwell Horton, Anish Prabhu, Anurag Ranjan, Ali Farhadi, Mohammad Rastegari
2021 arXiv   pre-print
Our models require no retraining, thus our subspace of models can be deployed entirely on-device to allow adaptive network compression at inference time.  ...  We present results for achieving arbitrarily fine-grained accuracy-efficiency trade-offs at inference time for structured and unstructured sparsity.  ...  Adaptive Compression for Structured Sparsity: Recent works train a single neural network which can be adaptively configured at inference time to execute at different compression levels.  ... 
arXiv:2110.04252v1 fatcat:u267be6z7ree5j4colsoowqdoa

Self-Taught Semi-Supervised Anomaly Detection on Upper Limb X-rays [article]

Antoine Spahr, Behzad Bozorgtabar, Jean-Philippe Thiran
2021 arXiv   pre-print
Supervised deep networks take for granted a large number of annotations by radiologists, which is often prohibitively very time-consuming to acquire.  ...  ., trained models suffer from overfitting to previously seen rare anomalies at training.  ...  architecture at inference time.  ... 
arXiv:2102.09895v2 fatcat:nzf4mmcqyrcutafxr4jhtobalu

Knowledge Aided Consistency for Weakly Supervised Phrase Grounding [article]

Kan Chen and Jiyang Gao and Ram Nevatia
2018 arXiv   pre-print
We propose a novel Knowledge Aided Consistency Network (KAC Net) which is optimized by reconstructing input query and proposal's information.  ...  Previous methods address this deficiency by training a grounding system via learning to reconstruct language information contained in input queries from predicted proposals.  ...  Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright notation thereon.  ... 
arXiv:1803.03879v1 fatcat:ddyuntrgxfg4zbsghbipjnjczq

Knowledge Aided Consistency for Weakly Supervised Phrase Grounding

Kan Chen, Jiyang Gao, Ram Nevatia
2018 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition  
We propose a novel Knowledge Aided Consistency Network (KAC Net) which is optimized by reconstructing input query and proposal's information.  ...  Previous methods address this deficiency by training a grounding system via learning to reconstruct language information contained in input queries from predicted proposals.  ...  Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright notation thereon.  ... 
doi:10.1109/cvpr.2018.00425 dblp:conf/cvpr/ChenGN18 fatcat:soi6pispmralrcv642wttutmgu

2020 Index IEEE Transactions on Very Large Scale Integration (VLSI) Systems Vol. 28

2020 IEEE Transactions on Very Large Scale Integration (vlsi) Systems  
., Conflux-An Asynchronous Two-to-One Multiplexor for Time-Division Multiplexing and Clockless, Tokenless Readout; TVLSI Feb. 2020 503-515 Holcomb, D., see 2685-2698 Holcomb, D.E., see 1807-1820 Homayoun  ...  ., +, TVLSI Dec. 2020 2612-2622 Compressed sensing A High-Throughput Subspace Pursuit Processor for ECG Recovery in Compressed Sensing Using Square-Root-Free MGS QR Decomposition.  ...  TiM-DNN: Ternary In-Memory Accelerator for Deep Neural Networks. Approximate Memory Compression.  ... 
doi:10.1109/tvlsi.2020.3041879 fatcat:33vb2eia2jfjpog4wei4peq5ge

Design of large polyphase filters in the Quadratic Residue Number System

Gian Carlo Cardarilli, Alberto Nannarelli, Yann Oster, Massimo Petricca, Marco Re
2010 2010 Conference Record of the Forty Fourth Asilomar Conference on Signals, Systems and Computers  
distributed inference problems in real-time.  ...  Each candidate is chosen to meet a specific subspace constraint to provide desirable constellation shaping for network coding at the relay.  ... 
doi:10.1109/acssc.2010.5757589 fatcat:ccxnu5owr5fyrcjcqukumerueq

A survey of sketches in traffic measurement: Design, Optimization, Application and Implementation [article]

Shangsen Li, Lailong Luo, Deke Guo, Qianzhen Zhang, Pengtao Fu
2021 arXiv   pre-print
At their cores, sketches usually maintain one or multiple counter array(s), and rely on hash functions to select the counter(s) for each flow.  ...  Finally,we highlight the open issues for future sketch-based network measurement research.  ...  Some studies have proposed compressing sketches when adapting to the available bandwidth before sending them in order to save space for time-adaptive updating. Compression for cardinality estimation.  ... 
arXiv:2012.07214v2 fatcat:lme2ghsshje3tag2m5q3xgvcna

Quantum Data Center: Theories and Applications [article]

Junyu Liu, Connor T. Hann, Liang Jiang
2022 arXiv   pre-print
QDC for distributed sensing through data compression.  ...  In this paper, we propose the Quantum Data Center (QDC), an architecture combining Quantum Random Access Memory (QRAM) and quantum networks.  ...  ACKNOWLEDGEMENT We thank Gideon Lee, John Preskill, Nicolas Sawaya, Xiaodi Wu, Xiao Yuan, Pei Zeng and Sisi Zhou for use-  ... 
arXiv:2207.14336v2 fatcat:aqrqdnrkbvdirk2i24qpz6ihm4

2021 Index IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Vol. 14

2021 IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing  
The Author Index contains the primary entry for each item, listed under the first author's name.  ...  Kang, L., +, JSTARS 2021 258-269 Mixed Compressive Sensing Back-Projection for SAR Focusing on Geo-Full Parameter Time Complexity (FPTC): A Method to Evaluate the Running Time of Machine Learning Classifiers  ...  Fang, B., +, JSTARS 2021 4946-4965 Data compression Adaptive Multisensor Acquisition via Spatial Contextual Information for Compressive Spectral Image Classification.  ... 
doi:10.1109/jstars.2022.3143012 fatcat:dnetkulbyvdyne7zxlblmek2qy

MARVEL - D3.1: Multimodal and privacy-aware audio-visual intelligence – initial version

Alexandros Iosifidis
2022 Zenodo  
as methodologies for improving the training and efficiency of AI models under supervised, unsupervised, and cross-modal contrastive learning settings.  ...  These include methods for Sound Event De- tection, Sound Event Localisation and Detection, Automated Audio Captioning, Visual Anomaly Detection, Visual Crowd Counting, Audio-Visual Crowd Counting, as well  ...  Such scalability features allow for extreme model compression and optimisation, while decoupling parameter count and computational cost in alignment with the harware-aware scaling paradigm.  ... 
doi:10.5281/zenodo.6821317 fatcat:eia7rkk5lfbg7khs3qcat5qd3m

2018 IndexIEEE Transactions on Very Large Scale Integration (VLSI) SystemsVol. 26

2018 IEEE Transactions on Very Large Scale Integration (vlsi) Systems  
., see 2723-2736 , VLSI Design of an ML-Based Power-Efficient Motion Estimation Controller for Intelligent Mobile Systems; TVLSI Feb. 2018 262-271 Hsieh, Y., see Tsai, Y., TVLSI May 2018 945-957  ...  Hsu, K., Chen, Y., Lee, Y., and Chang, S., Contactless Testing for Prebond Interposers; TVLSI June 2018 1005-1014 Hsu, Y., see Liu, Z., 1565-1574 Hu, J., see Wang, Y., TVLSI May 2018 805-817 Hu, J  ...  ., +, TVLSI Nov. 2018 2530-2541 Cellular neural networks Adaptive Precision Cellular Nonlinear Network.  ... 
doi:10.1109/tvlsi.2019.2892312 fatcat:rxiz5duc6jhdzjo4ybcxdajtbq

Localized Spectral Graph Filter Frames: A Unifying Framework, Survey of Design Considerations, and Numerical Comparison (Extended Cut) [article]

David I Shuman
2020 arXiv   pre-print
combination of building block signals can enable efficient and insightful visual or statistical analysis of the data, and such representations prove useful as regularizers in signal processing and machine learning  ...  Designing collections of building block signals -- or more formally, dictionaries of atoms -- that specifically account for the underlying graph structure as well as any available representative training  ...  ACKNOWLEDGMENTS The author would like to thank the anonymous reviewers and Hamid Behjat for constructive feedback on earlier versions of this article.  ... 
arXiv:2006.11220v2 fatcat:2fhnkgrlgfau7o4m2aoisoflju

The Minimum Description Length Principle for Pattern Mining: A Survey [article]

Esther Galbrun
2022 arXiv   pre-print
After giving an outline of relevant concepts from information theory and coding, as well as of work on the theory behind the MDL and similar principles, we review MDL-based methods for mining various types  ...  Proença and Jilles Vreeken for their comments on the first version of this document.  ...  Acknowledgments The author is grateful to Peggy Cellier for her feedback during the preparation and revisions of the manuscript, and to Hugo M.  ... 
arXiv:2007.14009v4 fatcat:jd5ab66jtjevro54docp6qdakm

Few-Shot Deep Adversarial Learning for Video-based Person Re-identification [article]

Lin Wu, Yang Wang, Hongzhi Yin, Meng Wang, Ling Shao
2019 arXiv   pre-print
Thus, matching videos for person re-ID demands flexible models to capture the dynamics in time-series observations and learn view-invariant representations with access to limited labeled training samples  ...  Also, it is noticed that learning effective video representations with view invariance is not explicitly addressed for which features exhibit different distributions otherwise.  ...  It should be at every time step or the last time step of a latent representation learning.  ... 
arXiv:1903.12395v2 fatcat:5iy2pgknqfhqjbr7baxkvwubdy

Ensembles of Random Sphere Cover Classifiers [article]

Anthony Bagnall, Reda Younsi
2014 arXiv   pre-print
We propose and evaluate alternative ensemble schemes for a new instance based learning classifier, the Randomised Sphere Cover (RSC) classifier.  ...  The randomised nature of RSC makes it ideal for use in ensembles.  ...  Ho: The random subspace method for con- margins for adaboost, Machine Learning, vol. 42, structing decision forests, IEEE Transactions on Pat- no. 3, 287-320, 2001.  ... 
arXiv:1409.4936v1 fatcat:gpvx737nhfcwdijs5zp3ouv3tu
« Previous Showing results 1 — 15 out of 241 results