This is an autogenerated issue. It uses the facebook/bart-large-cnn model to generate summaries for each papers appeared below.
Keyword: IoT (11 papers)
Toward Mixed Reality Hybrid Objects with IoT Avatar Agents
:brain: Authors: Alexis Morris, Jie Guan, Nadine Lessio, Yiyi Shao :paw_prints: Details:arXiv:2305.11960 :ramen: tl;dr: The internet-of-things (IoT) refers to the growing field of interconnected pervasive computing devices. This work addresses this by applying mixed reality toward expressing the underlying behaviors and states internal to IoT devices and IoT-enabled objects. This approach is a step toward new information presentation, interaction, and engagement capabilities for smart devices and environments.... :seedling: Read more →
LightESD: Fully-Automated and Lightweight Anomaly Detection Framework for Edge Computing
:brain: Authors: Ronit Das, Tie Luo :paw_prints: Details:arXiv:2305.12266 :ramen: tl;dr: Anomaly detection is widely used in a broad range of domains from cybersecurity to manufacturing, finance, and so on. We propose a fully-automated, lightweight, statistical learning based anomaly detection framework called LightESD. It is an on-device learning method without the need for data transfer between edge and server.... :sunflower: Comments: IEEE EDGE 2023, Chicago, USA, July 2023 :seedling: Read more →
Post-Quantum Hybrid Digital Signatures with Hardware-Support for Digital Twins
:brain: Authors: Saif E. Nouma, Attila A. Yavuz :paw_prints: Details:arXiv:2305.12298 :ramen: tl;dr: Digital Twins (DT) virtually model cyber-physical objects using Internet of Things (IoT) components. Trustworthiness of the streamed data is crucial which requires quantum safety and breach resiliency. NIST PQC signature standards are exorbitantly costly for low-end IoT without considering forward security.... :sunflower: Comments: 20 pages, 7 figures :seedling: Read more →
AgroTIC: Bridging the gap between farmers, agronomists, and merchants through smartphones and machine learning
:brain: Authors: Carlos Hinojosa, Karen Sanchez, Ariolfo Camacho, Henry Arguello :paw_prints: Details:arXiv:2305.12418 :ramen: tl;dr: AgroTIC is a smartphone-based application for agriculture that bridges the gap between farmers, agronomists, and merchants via ubiquitous technology and low-cost smartphones. It enables farmers to monitor their crop health with the assistance of agronmists, image processing, and deep learning. When farmers are ready to market their agricultural products, it provides a platform to connect them with merchants.... :seedling: Read more →
Routing for Intermittently-Powered Sensing Systems
:brain: Authors: Gaosheng Liu, Lin Wang :paw_prints: Details:arXiv:2305.12550 :ramen: tl;dr: Intermittent computing (IC) has received tremendous attention due to its high potential in perpetual sensing for Internet-of-Things (IoT) By harvesting ambient energy, battery-free devices can perform sensing intermittently without maintenance, thus significantly improving IoT sustainability. To build a practical intermittently-powered sensing system, efficient routing across battery- free devices for data delivery is essential.... :seedling: Read more →
More Perspectives Mean Better: Underwater Target Recognition and Localization with Multimodal Data via Symbiotic Transformer and Multiview Regression
:brain: Authors: Shipei Liu, Xiaoya Fan, Guowei Wu :paw_prints: Details:arXiv:2305.12701 :ramen: tl;dr: Underwater acoustic target recognition and localization (UATL) play important roles in marine exploration. The highly noisy acoustic signal and time-frequency interference among various sources pose big challenges to this task. We propose a multimodal approach to extract and fuse audio-visual-textual information to recognize and localize underwater targets through the Symbiotic Transformer and Multi-View Regression (MVR) method.... :seedling: Read more →
Spatiotemporal Attention-based Semantic Compression for Real-time Video Recognition
:brain: Authors: Nan Li, Mehdi Bennis, Alexandros Iosifidis, Qi Zhang :paw_prints: Details:arXiv:2305.12796 :ramen: tl;dr: This paper studies the computational offloading of video action recognition in edge computing. We propose a novel spatiotemporal attention-based autoencoder (STAE) architecture. ViT_STAE can compress the video dataset HMDB51 by 104x with only 5% accuracy loss, outperforming the state-of-the-art baseline DeepISC.... :sunflower: Comments: Submitted to IEEE Globecom 2023 :seedling: Read more →
Enhancing Next Active Object-based Egocentric Action Anticipation with Guided Attention
:brain: Authors: Sanket Thakur, Cigdem Beyan, Pietro Morerio, Vittorio Murino, Alessio Del Bue :paw_prints: Details:arXiv:2305.12953 :ramen: tl;dr: Short-term action anticipation (STA) in first-person videos is a challenging task that involves understanding the next active object interactions and predicting future actions. Existing action anticipation methods have primarily focused on utilizing features extracted from video clips, but often overlooked the importance of objects and their interactions. To this end, we propose a novel approach that applies a guided attention mechanism between the objects, and the spatiotemporal features.... :sunflower: Comments: Submitted to IEEE ICIP 2023 :seedling: Read more →
Contrastive Predictive Autoencoders for Dynamic Point Cloud Self-Supervised Learning
:brain: Authors: Xiaoxiao Sheng, Zhiqiang Shen, Gang Xiao :paw_prints: Details:arXiv:2305.12959 :ramen: tl;dr: We present a new self-supervised paradigm on point cloud sequence understanding. By combining classic contrast and reconstruction paradigms, it makes the learned representations with both global discrimination and local perception. We report the results on action recognition and gesture recognition under multiple experimental settings. The performances are comparable with supervised methods and show powerful transferability.... :sunflower: Comments: Accepted by AAAI2023 :seedling: Read more →
AudioToken: Adaptation of Text-Conditioned Diffusion Models for Audio-to-Image Generation
:brain: Authors: Guy Yariv, Itai Gat, Lior Wolf, Yossi Adi, Idan Schwartz :paw_prints: Details:arXiv:2305.13050 :ramen: tl;dr: In recent years, image generation has shown a great leap in performance, where diffusion models play a central role. Such models are mainly conditioned on textual descriptions. This begs the question "how can we adopt such models to be conditioned on other modalities?". In this paper, we propose a novel method utilizing latent diffusion models trained for text-to-image-generation to generate images conditioned on audio recordings.... :sunflower: Comments: Accepted to INTERSPEECH 2023 :seedling: Read more →
Real-life Implementation of Internet of Robotic Things Using 5 DoF Heterogeneous Robotic Arm
:brain: Authors: Sayed Erfan Arefin, Tasnia Ashrafi Heya, Jia Uddin :paw_prints: Details:arXiv:2305.13207 :ramen: tl;dr: Internet of Things (IoT) can be amalgamated with multitudinous objects such as thermostats, cars, lights, refrigerators, and many more appliances which will be able to build a connection via internet. Robotics can be another aspect which will get beneficial to be brought under the concept of IoT and is able to add a new perception in robotics having "Mechanical Smart Intelligence"... :seedling: Read more →
Keyword: edge computing (2 papers)
Soft Actor-Critic Learning-Based Joint Computing, Pushing, and Caching Framework in MEC Networks
:brain: Authors: Xiangyu Gao, Yaping Sun, Hao Chen, Xiaodong Xu, Shuguang Cui :paw_prints: Details:arXiv:2305.12099 :ramen: tl;dr: The mobile edge computing (MEC) network needs to be jointly optimized for computing, pushing, and caching. We propose a framework based on deep reinforcement learning that enables the dynamic orchestration of these three activities. The framework can implicitly predict user future requests using deep networks and push or cache the appropriate content to enhance performance.... :seedling: Read more →
Zero-shot Multi-level Feature Transmission Policy Powered by Semantic Knowledge Base
:brain: Authors: Yaping Sun, Hao Chen, Xiaodong Xu, Ping Zhang, Shuguang Cui :paw_prints: Details:arXiv:2305.12619 :ramen: tl;dr: Remote zero-shot object recognition has become a common and important task to solve for 6G. The optimization problem is a multi-choice knapsack problem, and thus very difficult to be optimized. This paper proposes an efficient algorithm based on convex concave procedure to find a high-quality solution.... :sunflower: Comments: 6 pages, 3 figures :seedling: Read more →
Keyword: Blockchain (2 papers)
Transparent and Traceable Food Supply Chain Management
:brain: Authors: Narayan Subramanian, Atharva Joshi, Daksh Bagga :paw_prints: Details:arXiv:2305.12188 :ramen: tl;dr: The food supply chain has a number of challenges, including a lack of transparency and disengagement among stakeholders. We propose a blockchain-based system for tracking a product's full path, from its raw components to the finished item in the store. Many advantages of the offered system include improved quality assessment, increased product transparency and traceability.... :seedling: Read more →
Network Participation and Accessibility of Proof-of-Stake (PoS) Blockchains: A Cross-platform Comparative Analysis
:brain: Authors: Jiseong Noh, Donghwan Kwon, Soohwan Cho, Neo C.K. Yiu :paw_prints: Details:arXiv:2305.13259 :ramen: tl;dr: A comparative analysis examined eleven Proof-of-Stake (PoS) consensus-based blockchain networks to assess their openness. The analysis suggested that networks with higher openness included Solana and Avalanche. BNB Chain, Klaytn, and Polygon measured with lower levels of openness.... :sunflower: Comments: 8 pages, 8 tables and 5 figures :seedling: Read more →
Keyword: Federated Learning (7 papers)
Can Public Large Language Models Help Private Cross-device Federated Learning?
:brain: Authors: Boxin Wang, Yibo Jacky Zhang, Yuan Cao, Bo Li, H. Brendan McMahan, Sewoong Oh, Zheng Xu, Manzil Zaheer :paw_prints: Details:arXiv:2305.12132 :ramen: tl;dr: We study (differentially) private federated learning (FL) of language models. The language models in cross-device FL are relatively small, which can be trained with meaningful formal user-level differential privacy. We propose a novel distribution matching algorithm with theoretical grounding to sample public data close to private data distribution.... :seedling: Read more →
Privacy in Multimodal Federated Human Activity Recognition
:brain: Authors: Alex Iacob, Pedro P. B. Gusmão, Nicholas D. Lane, Armand K. Koupai, Mohammud J. Bocus, Raúl Santos-Rodríguez, Robert J. Piechocki, Ryan McConville :paw_prints: Details:arXiv:2305.12134 :ramen: tl;dr: Human Activity Recognition (HAR) training data is often privacy-sensitive or held by non-cooperative entities. Federated Learning (FL) addresses such concerns by training ML models on edge clients. We show that the performance of FL for HAR depends on the assumed privacy level of the FL system.... :sunflower: Comments: In 3rd On-Device Intelligence Workshop at MLSys 2023, 8 pages :seedling: Read more →
One-Shot Federated Learning for LEO Constellations that Reduces Convergence Time from Days to 90 Minutes
:brain: Authors: Mohamed Elmahallawy, Tie Luo :paw_prints: Details:arXiv:2305.12316 :ramen: tl;dr: A Low Earth orbit (LEO) satellite constellation consists of a large number of small satellites traveling in space with high mobility. Machine learning can be utilized to analyze these mobility data to address global challenges. Federated Learning (FL) is a promising approach because it eliminates the need for transmitting raw data.... :sunflower: Comments: This article belongs to The 24th IEEE International Conference on Mobile Data Management (MDM 2023) :seedling: Read more →
Confidence-aware Personalized Federated Learning via Variational Expectation Maximization
:brain: Authors: Junyi Zhu, Xingchen Ma, Matthew B. Blaschko :paw_prints: Details:arXiv:2305.12557 :ramen: tl;dr: Personalized Federated Learning (PFL) attempts to solve this challenge via locally adapted models. A global model is introduced as a latent variable to augment the joint distribution of clients' parameters and capture the common trends of different clients. Our algorithm gives rise to a closed-form estimation of a confidence value.... :sunflower: Comments: Accepted at CVPR 2023 :seedling: Read more →
FSSA: Efficient 3-Round Secure Aggregation for Privacy-Preserving Federated Learning
:brain: Authors: Fucai Luo, Saif Al-Kuwari, Haiyan Wang, Xingfu Yan :paw_prints: Details:arXiv:2305.12950 :ramen: tl;dr: Federated learning (FL) allows a large number of clients to collaboratively train machine learning (ML) models. Recent attacks on FL demonstrate that local gradients may leak information about local training data. Bonawitz (CCS 2017) proposed a secure aggregation protocol that allows a server to compute the sum of clients' localGradients in a secure manner. However, their secure aggregation Protocol requires at least 4 rounds of communication between each client and the server in each training iteration.... :seedling: Read more →
Federated Learning of Medical Concepts Embedding using BEHRT
:brain: Authors: Ofir Ben Shoham, Nadav Rappoport :paw_prints: Details:arXiv:2305.13052 :ramen: tl;dr: Federated Learning (FL) is an algorithmic approach that allows learning a shared model using data in multiple locations without the need to store all data in a central place. The approach is based on an embedding model like BEHRT, a deep neural sequence transduction model for EHR.... :seedling: Read more →
Explicit Personalization and Local Training: Double Communication Acceleration in Federated Learning
:brain: Authors: Kai Yi, Laurent Condat, Peter Richtárik :paw_prints: Details:arXiv:2305.13170 :ramen: tl;dr: Federated Learning is an evolving machine learning paradigm. Multiple clients perform computations based on their individual private data. A common strategy to curtail communication costs is Local Training. We introduce Scafflix, a novel algorithm that efficiently integrates explicit personalization with local training.... :seedling: Read more →
This is an autogenerated issue. It uses the facebook/bart-large-cnn model to generate summaries for each papers appeared below.
Keyword: IoT (11 papers)
Toward Mixed Reality Hybrid Objects with IoT Avatar Agents
:brain: Authors: Alexis Morris, Jie Guan, Nadine Lessio, Yiyi Shao:paw_prints: Details: arXiv:2305.11960
:ramen: tl;dr: The internet-of-things (IoT) refers to the growing field of interconnected pervasive computing devices. This work addresses this by applying mixed reality toward expressing the underlying behaviors and states internal to IoT devices and IoT-enabled objects. This approach is a step toward new information presentation, interaction, and engagement capabilities for smart devices and environments....
:seedling: Read more →
LightESD: Fully-Automated and Lightweight Anomaly Detection Framework for Edge Computing
:brain: Authors: Ronit Das, Tie Luo:paw_prints: Details: arXiv:2305.12266
:ramen: tl;dr: Anomaly detection is widely used in a broad range of domains from cybersecurity to manufacturing, finance, and so on. We propose a fully-automated, lightweight, statistical learning based anomaly detection framework called LightESD. It is an on-device learning method without the need for data transfer between edge and server....
:sunflower: Comments: IEEE EDGE 2023, Chicago, USA, July 2023
:seedling: Read more →
Post-Quantum Hybrid Digital Signatures with Hardware-Support for Digital Twins
:brain: Authors: Saif E. Nouma, Attila A. Yavuz:paw_prints: Details: arXiv:2305.12298
:ramen: tl;dr: Digital Twins (DT) virtually model cyber-physical objects using Internet of Things (IoT) components. Trustworthiness of the streamed data is crucial which requires quantum safety and breach resiliency. NIST PQC signature standards are exorbitantly costly for low-end IoT without considering forward security....
:sunflower: Comments: 20 pages, 7 figures
:seedling: Read more →
AgroTIC: Bridging the gap between farmers, agronomists, and merchants through smartphones and machine learning
:brain: Authors: Carlos Hinojosa, Karen Sanchez, Ariolfo Camacho, Henry Arguello:paw_prints: Details: arXiv:2305.12418
:ramen: tl;dr: AgroTIC is a smartphone-based application for agriculture that bridges the gap between farmers, agronomists, and merchants via ubiquitous technology and low-cost smartphones. It enables farmers to monitor their crop health with the assistance of agronmists, image processing, and deep learning. When farmers are ready to market their agricultural products, it provides a platform to connect them with merchants....
:seedling: Read more →
Routing for Intermittently-Powered Sensing Systems
:brain: Authors: Gaosheng Liu, Lin Wang:paw_prints: Details: arXiv:2305.12550
:ramen: tl;dr: Intermittent computing (IC) has received tremendous attention due to its high potential in perpetual sensing for Internet-of-Things (IoT) By harvesting ambient energy, battery-free devices can perform sensing intermittently without maintenance, thus significantly improving IoT sustainability. To build a practical intermittently-powered sensing system, efficient routing across battery- free devices for data delivery is essential....
:seedling: Read more →
More Perspectives Mean Better: Underwater Target Recognition and Localization with Multimodal Data via Symbiotic Transformer and Multiview Regression
:brain: Authors: Shipei Liu, Xiaoya Fan, Guowei Wu:paw_prints: Details: arXiv:2305.12701
:ramen: tl;dr: Underwater acoustic target recognition and localization (UATL) play important roles in marine exploration. The highly noisy acoustic signal and time-frequency interference among various sources pose big challenges to this task. We propose a multimodal approach to extract and fuse audio-visual-textual information to recognize and localize underwater targets through the Symbiotic Transformer and Multi-View Regression (MVR) method....
:seedling: Read more →
Spatiotemporal Attention-based Semantic Compression for Real-time Video Recognition
:brain: Authors: Nan Li, Mehdi Bennis, Alexandros Iosifidis, Qi Zhang:paw_prints: Details: arXiv:2305.12796
:ramen: tl;dr: This paper studies the computational offloading of video action recognition in edge computing. We propose a novel spatiotemporal attention-based autoencoder (STAE) architecture. ViT_STAE can compress the video dataset HMDB51 by 104x with only 5% accuracy loss, outperforming the state-of-the-art baseline DeepISC....
:sunflower: Comments: Submitted to IEEE Globecom 2023
:seedling: Read more →
Enhancing Next Active Object-based Egocentric Action Anticipation with Guided Attention
:brain: Authors: Sanket Thakur, Cigdem Beyan, Pietro Morerio, Vittorio Murino, Alessio Del Bue:paw_prints: Details: arXiv:2305.12953
:ramen: tl;dr: Short-term action anticipation (STA) in first-person videos is a challenging task that involves understanding the next active object interactions and predicting future actions. Existing action anticipation methods have primarily focused on utilizing features extracted from video clips, but often overlooked the importance of objects and their interactions. To this end, we propose a novel approach that applies a guided attention mechanism between the objects, and the spatiotemporal features....
:sunflower: Comments: Submitted to IEEE ICIP 2023
:seedling: Read more →
Contrastive Predictive Autoencoders for Dynamic Point Cloud Self-Supervised Learning
:brain: Authors: Xiaoxiao Sheng, Zhiqiang Shen, Gang Xiao:paw_prints: Details: arXiv:2305.12959
:ramen: tl;dr: We present a new self-supervised paradigm on point cloud sequence understanding. By combining classic contrast and reconstruction paradigms, it makes the learned representations with both global discrimination and local perception. We report the results on action recognition and gesture recognition under multiple experimental settings. The performances are comparable with supervised methods and show powerful transferability....
:sunflower: Comments: Accepted by AAAI2023
:seedling: Read more →
AudioToken: Adaptation of Text-Conditioned Diffusion Models for Audio-to-Image Generation
:brain: Authors: Guy Yariv, Itai Gat, Lior Wolf, Yossi Adi, Idan Schwartz:paw_prints: Details: arXiv:2305.13050
:ramen: tl;dr: In recent years, image generation has shown a great leap in performance, where diffusion models play a central role. Such models are mainly conditioned on textual descriptions. This begs the question "how can we adopt such models to be conditioned on other modalities?". In this paper, we propose a novel method utilizing latent diffusion models trained for text-to-image-generation to generate images conditioned on audio recordings....
:sunflower: Comments: Accepted to INTERSPEECH 2023
:seedling: Read more →
Real-life Implementation of Internet of Robotic Things Using 5 DoF Heterogeneous Robotic Arm
:brain: Authors: Sayed Erfan Arefin, Tasnia Ashrafi Heya, Jia Uddin:paw_prints: Details: arXiv:2305.13207
:ramen: tl;dr: Internet of Things (IoT) can be amalgamated with multitudinous objects such as thermostats, cars, lights, refrigerators, and many more appliances which will be able to build a connection via internet. Robotics can be another aspect which will get beneficial to be brought under the concept of IoT and is able to add a new perception in robotics having "Mechanical Smart Intelligence"...
:seedling: Read more →
Keyword: edge computing (2 papers)
Soft Actor-Critic Learning-Based Joint Computing, Pushing, and Caching Framework in MEC Networks
:brain: Authors: Xiangyu Gao, Yaping Sun, Hao Chen, Xiaodong Xu, Shuguang Cui:paw_prints: Details: arXiv:2305.12099
:ramen: tl;dr: The mobile edge computing (MEC) network needs to be jointly optimized for computing, pushing, and caching. We propose a framework based on deep reinforcement learning that enables the dynamic orchestration of these three activities. The framework can implicitly predict user future requests using deep networks and push or cache the appropriate content to enhance performance....
:seedling: Read more →
Zero-shot Multi-level Feature Transmission Policy Powered by Semantic Knowledge Base
:brain: Authors: Yaping Sun, Hao Chen, Xiaodong Xu, Ping Zhang, Shuguang Cui:paw_prints: Details: arXiv:2305.12619
:ramen: tl;dr: Remote zero-shot object recognition has become a common and important task to solve for 6G. The optimization problem is a multi-choice knapsack problem, and thus very difficult to be optimized. This paper proposes an efficient algorithm based on convex concave procedure to find a high-quality solution....
:sunflower: Comments: 6 pages, 3 figures
:seedling: Read more →
Keyword: Blockchain (2 papers)
Transparent and Traceable Food Supply Chain Management
:brain: Authors: Narayan Subramanian, Atharva Joshi, Daksh Bagga:paw_prints: Details: arXiv:2305.12188
:ramen: tl;dr: The food supply chain has a number of challenges, including a lack of transparency and disengagement among stakeholders. We propose a blockchain-based system for tracking a product's full path, from its raw components to the finished item in the store. Many advantages of the offered system include improved quality assessment, increased product transparency and traceability....
:seedling: Read more →
Network Participation and Accessibility of Proof-of-Stake (PoS) Blockchains: A Cross-platform Comparative Analysis
:brain: Authors: Jiseong Noh, Donghwan Kwon, Soohwan Cho, Neo C.K. Yiu:paw_prints: Details: arXiv:2305.13259
:ramen: tl;dr: A comparative analysis examined eleven Proof-of-Stake (PoS) consensus-based blockchain networks to assess their openness. The analysis suggested that networks with higher openness included Solana and Avalanche. BNB Chain, Klaytn, and Polygon measured with lower levels of openness....
:sunflower: Comments: 8 pages, 8 tables and 5 figures
:seedling: Read more →
Keyword: Federated Learning (7 papers)
Can Public Large Language Models Help Private Cross-device Federated Learning?
:brain: Authors: Boxin Wang, Yibo Jacky Zhang, Yuan Cao, Bo Li, H. Brendan McMahan, Sewoong Oh, Zheng Xu, Manzil Zaheer:paw_prints: Details: arXiv:2305.12132
:ramen: tl;dr: We study (differentially) private federated learning (FL) of language models. The language models in cross-device FL are relatively small, which can be trained with meaningful formal user-level differential privacy. We propose a novel distribution matching algorithm with theoretical grounding to sample public data close to private data distribution....
:seedling: Read more →
Privacy in Multimodal Federated Human Activity Recognition
:brain: Authors: Alex Iacob, Pedro P. B. Gusmão, Nicholas D. Lane, Armand K. Koupai, Mohammud J. Bocus, Raúl Santos-Rodríguez, Robert J. Piechocki, Ryan McConville:paw_prints: Details: arXiv:2305.12134
:ramen: tl;dr: Human Activity Recognition (HAR) training data is often privacy-sensitive or held by non-cooperative entities. Federated Learning (FL) addresses such concerns by training ML models on edge clients. We show that the performance of FL for HAR depends on the assumed privacy level of the FL system....
:sunflower: Comments: In 3rd On-Device Intelligence Workshop at MLSys 2023, 8 pages
:seedling: Read more →
One-Shot Federated Learning for LEO Constellations that Reduces Convergence Time from Days to 90 Minutes
:brain: Authors: Mohamed Elmahallawy, Tie Luo:paw_prints: Details: arXiv:2305.12316
:ramen: tl;dr: A Low Earth orbit (LEO) satellite constellation consists of a large number of small satellites traveling in space with high mobility. Machine learning can be utilized to analyze these mobility data to address global challenges. Federated Learning (FL) is a promising approach because it eliminates the need for transmitting raw data....
:sunflower: Comments: This article belongs to The 24th IEEE International Conference on Mobile Data Management (MDM 2023)
:seedling: Read more →
Confidence-aware Personalized Federated Learning via Variational Expectation Maximization
:brain: Authors: Junyi Zhu, Xingchen Ma, Matthew B. Blaschko:paw_prints: Details: arXiv:2305.12557
:ramen: tl;dr: Personalized Federated Learning (PFL) attempts to solve this challenge via locally adapted models. A global model is introduced as a latent variable to augment the joint distribution of clients' parameters and capture the common trends of different clients. Our algorithm gives rise to a closed-form estimation of a confidence value....
:sunflower: Comments: Accepted at CVPR 2023
:seedling: Read more →
FSSA: Efficient 3-Round Secure Aggregation for Privacy-Preserving Federated Learning
:brain: Authors: Fucai Luo, Saif Al-Kuwari, Haiyan Wang, Xingfu Yan:paw_prints: Details: arXiv:2305.12950
:ramen: tl;dr: Federated learning (FL) allows a large number of clients to collaboratively train machine learning (ML) models. Recent attacks on FL demonstrate that local gradients may leak information about local training data. Bonawitz (CCS 2017) proposed a secure aggregation protocol that allows a server to compute the sum of clients' localGradients in a secure manner. However, their secure aggregation Protocol requires at least 4 rounds of communication between each client and the server in each training iteration....
:seedling: Read more →
Federated Learning of Medical Concepts Embedding using BEHRT
:brain: Authors: Ofir Ben Shoham, Nadav Rappoport:paw_prints: Details: arXiv:2305.13052
:ramen: tl;dr: Federated Learning (FL) is an algorithmic approach that allows learning a shared model using data in multiple locations without the need to store all data in a central place. The approach is based on an embedding model like BEHRT, a deep neural sequence transduction model for EHR....
:seedling: Read more →
Explicit Personalization and Local Training: Double Communication Acceleration in Federated Learning
:brain: Authors: Kai Yi, Laurent Condat, Peter Richtárik:paw_prints: Details: arXiv:2305.13170
:ramen: tl;dr: Federated Learning is an evolving machine learning paradigm. Multiple clients perform computations based on their individual private data. A common strategy to curtail communication costs is Local Training. We introduce Scafflix, a novel algorithm that efficiently integrates explicit personalization with local training....
:seedling: Read more →