Md. Farhan Tasnim Oshim, J. Killingback, Dave Follette, Huaishu Peng, Tauhidur Rahman
Knowing how and when people interact with their surroundings is crucial for constructing dynamic and intelligent environments. Despite the importance of this problem, an attainable and simple solution is still lacking. Current solutions often require powered sensors on monitored objects or users themselves. Many such systems use batteries [1-3], which are costly and time consuming to replace. Some powered systems connect to the grid, which may save swapping batteries, but at the price of restricted placement options. Other solutions use passive tags on monitored objects or require no tags at all, but many of these systems have prohibitive characteristics. For instance, camera-based systems [4,5] generally will not work if their view is occluded. Many other systems that rely on passive tags or do not use tags require direct line-of-sight or close proximity to work. As such, our goal was to design and develop small, cheap, easy-to-install tags that do not require any batteries, silicon chips or discrete electronic components, which can be monitored without direct line-of-sight.
{"title":"Building MechanoBeat","authors":"Md. Farhan Tasnim Oshim, J. Killingback, Dave Follette, Huaishu Peng, Tauhidur Rahman","doi":"10.1145/3583571.3583573","DOIUrl":"https://doi.org/10.1145/3583571.3583573","url":null,"abstract":"Knowing how and when people interact with their surroundings is crucial for constructing dynamic and intelligent environments. Despite the importance of this problem, an attainable and simple solution is still lacking. Current solutions often require powered sensors on monitored objects or users themselves. Many such systems use batteries [1-3], which are costly and time consuming to replace. Some powered systems connect to the grid, which may save swapping batteries, but at the price of restricted placement options. Other solutions use passive tags on monitored objects or require no tags at all, but many of these systems have prohibitive characteristics. For instance, camera-based systems [4,5] generally will not work if their view is occluded. Many other systems that rely on passive tags or do not use tags require direct line-of-sight or close proximity to work. As such, our goal was to design and develop small, cheap, easy-to-install tags that do not require any batteries, silicon chips or discrete electronic components, which can be monitored without direct line-of-sight.","PeriodicalId":29918,"journal":{"name":"GetMobile-Mobile Computing & Communications Review","volume":"11 1","pages":"5 - 13"},"PeriodicalIF":1.0,"publicationDate":"2023-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73544830","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Tusher Chakraborty, Heping Shi, Zerina Kapetanovic, B. Priyantha, Deepak Vasisht, Binh Vu, Parag Pandit, Prasad Pillai, Y. Chabria, Andrew Nelson, Michael Daum, Ranveer Chandra
The epoch-making proliferation of Internet of Things (IoT) networks in recent years has brought connectivity to homes, cities, farms, and many other industries. ISM bands are accommodating most of these networks around the world. However, our experience from several global deployments has shown that such networks are bottlenecked by communication range and bandwidth. With these IoT deployment constraints in mind, we propose a new connectivity solution, Whisper [1], where IoT devices can opportunistically transmit data in the TV White Space (TVWS) spectrum, while protecting incumbents from receiving harmful interference.
{"title":"Whisper","authors":"Tusher Chakraborty, Heping Shi, Zerina Kapetanovic, B. Priyantha, Deepak Vasisht, Binh Vu, Parag Pandit, Prasad Pillai, Y. Chabria, Andrew Nelson, Michael Daum, Ranveer Chandra","doi":"10.1145/3583571.3583580","DOIUrl":"https://doi.org/10.1145/3583571.3583580","url":null,"abstract":"The epoch-making proliferation of Internet of Things (IoT) networks in recent years has brought connectivity to homes, cities, farms, and many other industries. ISM bands are accommodating most of these networks around the world. However, our experience from several global deployments has shown that such networks are bottlenecked by communication range and bandwidth. With these IoT deployment constraints in mind, we propose a new connectivity solution, Whisper [1], where IoT devices can opportunistically transmit data in the TV White Space (TVWS) spectrum, while protecting incumbents from receiving harmful interference.","PeriodicalId":29918,"journal":{"name":"GetMobile-Mobile Computing & Communications Review","volume":"9 1","pages":"32 - 35"},"PeriodicalIF":1.0,"publicationDate":"2023-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74029278","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With the advancement of the Internet of Things (IoT), billions of devices will be connected to the Internet, enabling new applications such as digital twin, augmented reality, and smart home. These applications have placed a huge strain on today's wireless network. mmWave technology is promising to solve this problem by providing a large bandwidth over the very-high-frequency spectrum band. However, most mmWave radios and platforms have much higher power consumption than what IoT devices and their applications can afford. Hence, mmWave networks cannot be utilized in most IoT applications today. In this work, we present a novel low-power mmWave platform, which brings this technology to IoT applications. Our approach to design this platform is to take a holistic view and optimize the whole wireless system by considering practical challenges in mmWave communication. Our lowcost and low-power platform not only brings mmWave communication to IoT applications, but also enables researchers that do not have hardware background to work on mmWave research.
{"title":"A Low-Power mmWave Platform for the Internet of Things","authors":"M. Mazaheri, Omid Salehi-Abari","doi":"10.1145/3583571.3583575","DOIUrl":"https://doi.org/10.1145/3583571.3583575","url":null,"abstract":"With the advancement of the Internet of Things (IoT), billions of devices will be connected to the Internet, enabling new applications such as digital twin, augmented reality, and smart home. These applications have placed a huge strain on today's wireless network. mmWave technology is promising to solve this problem by providing a large bandwidth over the very-high-frequency spectrum band. However, most mmWave radios and platforms have much higher power consumption than what IoT devices and their applications can afford. Hence, mmWave networks cannot be utilized in most IoT applications today. In this work, we present a novel low-power mmWave platform, which brings this technology to IoT applications. Our approach to design this platform is to take a holistic view and optimize the whole wireless system by considering practical challenges in mmWave communication. Our lowcost and low-power platform not only brings mmWave communication to IoT applications, but also enables researchers that do not have hardware background to work on mmWave research.","PeriodicalId":29918,"journal":{"name":"GetMobile-Mobile Computing & Communications Review","volume":"13 1","pages":"14 - 18"},"PeriodicalIF":1.0,"publicationDate":"2023-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89574267","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Junbo Zhang, Elahe Soltanaghai, Artur Balanuta, Reese Grimsley, Swarun Kumar, Anthony G. Rowe
Can we read ultra-low-power sensors in a large industrial or commercial building with a single reader using the power line system? As the manufacturing industry becomes more and more automated, IoT sensors are also being widely deployed inside industrial buildings. Given the significant cost associated with retrofitting an industrial building, a wired network for IoT installation might not be desirable. On the other hand, long-range wireless networks are either power-hungry (e.g., Wi-Fi or cellular), or support only a low data rate (e.g., LoRa). In this paper, we explore an alternative approach: leveraging the power line infrastructure to enable building-scale wireless backscatter communication.
{"title":"PLatter","authors":"Junbo Zhang, Elahe Soltanaghai, Artur Balanuta, Reese Grimsley, Swarun Kumar, Anthony G. Rowe","doi":"10.1145/3583571.3583577","DOIUrl":"https://doi.org/10.1145/3583571.3583577","url":null,"abstract":"Can we read ultra-low-power sensors in a large industrial or commercial building with a single reader using the power line system? As the manufacturing industry becomes more and more automated, IoT sensors are also being widely deployed inside industrial buildings. Given the significant cost associated with retrofitting an industrial building, a wired network for IoT installation might not be desirable. On the other hand, long-range wireless networks are either power-hungry (e.g., Wi-Fi or cellular), or support only a low data rate (e.g., LoRa). In this paper, we explore an alternative approach: leveraging the power line infrastructure to enable building-scale wireless backscatter communication.","PeriodicalId":29918,"journal":{"name":"GetMobile-Mobile Computing & Communications Review","volume":"220 1","pages":"19 - 22"},"PeriodicalIF":1.0,"publicationDate":"2023-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75891595","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jayanth Shenoy, Zikun Liu, Bill Tao, Zachary Kabelac, Deepak Vasisht
In the last decade, both academia and industry have relied on FMCWradar based radio-frequency (RF) sensors to enable through-wall human tracking. These sensors capture reflections from human bodies to track occupancy of rooms [1], motion patterns of occupants [1,2], their daily activities [3], and their health metrics [4, 5]. Recently, Google has incorporated high frequency FMCW-based sensing into their smart home devices [6, 7], and Amazon received an FCC waiver [8] to conduct testing for the same.
{"title":"RF-Protect","authors":"Jayanth Shenoy, Zikun Liu, Bill Tao, Zachary Kabelac, Deepak Vasisht","doi":"10.1145/3583571.3583579","DOIUrl":"https://doi.org/10.1145/3583571.3583579","url":null,"abstract":"In the last decade, both academia and industry have relied on FMCWradar based radio-frequency (RF) sensors to enable through-wall human tracking. These sensors capture reflections from human bodies to track occupancy of rooms [1], motion patterns of occupants [1,2], their daily activities [3], and their health metrics [4, 5]. Recently, Google has incorporated high frequency FMCW-based sensing into their smart home devices [6, 7], and Amazon received an FCC waiver [8] to conduct testing for the same.","PeriodicalId":29918,"journal":{"name":"GetMobile-Mobile Computing & Communications Review","volume":"11 1","pages":"28 - 31"},"PeriodicalIF":1.0,"publicationDate":"2023-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72726976","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The state of underwater networking today is similar to ARPANET back in the 1970s, when only a select few people with expensive hardware resources had access to the technology. We present the first acoustic system that brings underwater networking capabilities to existing mobile devices like smartphones and smart watches. Our software-only solution leverages audio sensors, i.e., microphones and speakers, ubiquitous in today's devices, to enable acoustic underwater communication between mobile devices. To achieve this, we design a communication system that adapts in real-time to differences in frequency responses across mobile devices, changes in multipath and noise levels at different locations and dynamic channel changes due to mobility. We evaluate our system in six different real-world underwater environments in the presence of boats, ships and people fishing and kayaking. With the release of Apple Watch Ultra that is specifically designed for underwater settings, our software-based approach has the potential to democratize underwater networking capabilities by making them widely available to anyone with a mobile device.
{"title":"Bringing Underwater Networking to the 21st Century","authors":"Justin Chan, Tuochao Chen, Shyamnath Gollakota","doi":"10.1145/3583571.3583578","DOIUrl":"https://doi.org/10.1145/3583571.3583578","url":null,"abstract":"The state of underwater networking today is similar to ARPANET back in the 1970s, when only a select few people with expensive hardware resources had access to the technology. We present the first acoustic system that brings underwater networking capabilities to existing mobile devices like smartphones and smart watches. Our software-only solution leverages audio sensors, i.e., microphones and speakers, ubiquitous in today's devices, to enable acoustic underwater communication between mobile devices. To achieve this, we design a communication system that adapts in real-time to differences in frequency responses across mobile devices, changes in multipath and noise levels at different locations and dynamic channel changes due to mobility. We evaluate our system in six different real-world underwater environments in the presence of boats, ships and people fishing and kayaking. With the release of Apple Watch Ultra that is specifically designed for underwater settings, our software-based approach has the potential to democratize underwater networking capabilities by making them widely available to anyone with a mobile device.","PeriodicalId":29918,"journal":{"name":"GetMobile-Mobile Computing & Communications Review","volume":"3 1","pages":"23 - 27"},"PeriodicalIF":1.0,"publicationDate":"2023-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80207083","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Colleen Josephson, W. Shuai, Gabriela Marcano, P. Pannuto, Josiah D. Hester, George Wells
The emergence of the Internet of Things and pervasive sensor networks have generated a surge of research in energy scavenging techniques. We know well that harvesting RF, solar, or kinetic energy enables the creation of battery-free devices that can be used where frequent battery changes or dedicated power lines are impractical. One unusual yet ubiquitous source of power is soil (earth itself) - or more accurately, bacterial communities in soil. Microbial fuel cells (MFCs) are electrochemical cells that harness the activities of microbes that naturally occur in soil, wetlands, and wastewater. MFCs have been a topic of research in environmental engineering and microbiology for decades, but are a relatively new topic in electronics design and research. Most low-power electronics have traditionally opted for batteries, RF energy, or solar cells. This is changing, however, as the limitations and costs of these energy sources hamper our ability to deploy useful systems that last for decades in challenging environments. If large-scale, long-term applications like underground infrastructure monitoring, smart farming, and sensing for conservation are to be possible, we must rethink the energy source.
物联网(Internet of Things)和无处不在的传感器网络的出现，催生了能量收集技术的研究热潮。我们很清楚，收集射频、太阳能或动能可以创造出无电池设备，这些设备可以在频繁更换电池或专用电源线不切实际的地方使用。一种不寻常但却无处不在的能源是土壤(土壤本身)——或者更准确地说，是土壤中的细菌群落。微生物燃料电池(mfc)是利用土壤、湿地和废水中自然存在的微生物活动的电化学电池。几十年来，mfc一直是环境工程和微生物学领域的研究课题，但在电子设计和研究中却是一个相对较新的课题。大多数低功耗电子产品传统上都选择电池、射频能量或太阳能电池。然而，这种情况正在改变，因为这些能源的局限性和成本阻碍了我们在具有挑战性的环境中部署持续数十年的有用系统的能力。如果地下基础设施监测、智能农业和保护传感等大规模长期应用成为可能，我们必须重新思考能源。
{"title":"The Future of Clean Computing May Be Dirty","authors":"Colleen Josephson, W. Shuai, Gabriela Marcano, P. Pannuto, Josiah D. Hester, George Wells","doi":"10.1145/3568113.3568117","DOIUrl":"https://doi.org/10.1145/3568113.3568117","url":null,"abstract":"The emergence of the Internet of Things and pervasive sensor networks have generated a surge of research in energy scavenging techniques. We know well that harvesting RF, solar, or kinetic energy enables the creation of battery-free devices that can be used where frequent battery changes or dedicated power lines are impractical. One unusual yet ubiquitous source of power is soil (earth itself) - or more accurately, bacterial communities in soil. Microbial fuel cells (MFCs) are electrochemical cells that harness the activities of microbes that naturally occur in soil, wetlands, and wastewater. MFCs have been a topic of research in environmental engineering and microbiology for decades, but are a relatively new topic in electronics design and research. Most low-power electronics have traditionally opted for batteries, RF energy, or solar cells. This is changing, however, as the limitations and costs of these energy sources hamper our ability to deploy useful systems that last for decades in challenging environments. If large-scale, long-term applications like underground infrastructure monitoring, smart farming, and sensing for conservation are to be possible, we must rethink the energy source.","PeriodicalId":29918,"journal":{"name":"GetMobile-Mobile Computing & Communications Review","volume":"11 1","pages":"9 - 15"},"PeriodicalIF":1.0,"publicationDate":"2022-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84299017","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Huatao Xu, Pengfei Zhou, R. Tan, Mo Li, Guobin Shen
Deep learning greatly empowers Inertial Measurement Unit (IMU) sensors for a wide range of sensing applications. Most existing works require substantial amounts of wellcurated labeled data to train IMU-based sensing models, which incurs high annotation and training costs. Compared with labeled data, unlabeled IMU data are abundant and easily accessible. This article presents a novel representation learning model that can make use of unlabeled IMU data and extract generalized rather than task-specific features. With the representations learned via our model, task-specific models trained with limited labeled samples can achieve superior performances in typical IMU sensing applications, such as Human Activity Recognition (HAR).
{"title":"LIMU-BERT","authors":"Huatao Xu, Pengfei Zhou, R. Tan, Mo Li, Guobin Shen","doi":"10.1145/3568113.3568124","DOIUrl":"https://doi.org/10.1145/3568113.3568124","url":null,"abstract":"Deep learning greatly empowers Inertial Measurement Unit (IMU) sensors for a wide range of sensing applications. Most existing works require substantial amounts of wellcurated labeled data to train IMU-based sensing models, which incurs high annotation and training costs. Compared with labeled data, unlabeled IMU data are abundant and easily accessible. This article presents a novel representation learning model that can make use of unlabeled IMU data and extract generalized rather than task-specific features. With the representations learned via our model, task-specific models trained with limited labeled samples can achieve superior performances in typical IMU sensing applications, such as Human Activity Recognition (HAR).","PeriodicalId":29918,"journal":{"name":"GetMobile-Mobile Computing & Communications Review","volume":"10 1","pages":"39 - 42"},"PeriodicalIF":1.0,"publicationDate":"2022-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88680356","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Xiaoran Fan, Longfei Shangguan, Siddharth Rupavatharam, Yanyong Zhang, Jie Xiong, Yunfei Ma, R. Howard
Headphones continue to grow more intelligent as new functions (e.g., touch-based gesture control) appear. These functions usually rely on auxiliary sensors (e.g., accelerometer and gyroscope) that are available in smart headphones. However, for those headphones that do not have such sensors, supporting these functions becomes a daunting task. This paper presents HeadFi, a new design paradigm for bringing intelligence to all headphones. Instead of adding auxiliary sensors into headphones, HeadFi turns the pair of drivers that are readily available inside all headphones into a versatile sensor to enable new applications, spanning across mobile health, user-interface, and context-awareness. HeadFi works as a plug-in peripheral connecting the headphones and the pairing device (e.g., a smartphone). The simplicity (can be as simple as just two resistors) and small form factor of this design lend itself to be embedded into the pairing device as an integrated circuit. We envision that HeadFi can serve as a vital supplementary solution to existing smart headphone design by directly transforming large amounts of existing "dumb" headphones into intelligent ones.
{"title":"A New Design Paradigm for Enabling Smart Headphonse","authors":"Xiaoran Fan, Longfei Shangguan, Siddharth Rupavatharam, Yanyong Zhang, Jie Xiong, Yunfei Ma, R. Howard","doi":"10.1145/3568113.3568122","DOIUrl":"https://doi.org/10.1145/3568113.3568122","url":null,"abstract":"Headphones continue to grow more intelligent as new functions (e.g., touch-based gesture control) appear. These functions usually rely on auxiliary sensors (e.g., accelerometer and gyroscope) that are available in smart headphones. However, for those headphones that do not have such sensors, supporting these functions becomes a daunting task. This paper presents HeadFi, a new design paradigm for bringing intelligence to all headphones. Instead of adding auxiliary sensors into headphones, HeadFi turns the pair of drivers that are readily available inside all headphones into a versatile sensor to enable new applications, spanning across mobile health, user-interface, and context-awareness. HeadFi works as a plug-in peripheral connecting the headphones and the pairing device (e.g., a smartphone). The simplicity (can be as simple as just two resistors) and small form factor of this design lend itself to be embedded into the pairing device as an integrated circuit. We envision that HeadFi can serve as a vital supplementary solution to existing smart headphone design by directly transforming large amounts of existing \"dumb\" headphones into intelligent ones.","PeriodicalId":29918,"journal":{"name":"GetMobile-Mobile Computing & Communications Review","volume":"36 1","pages":"27 - 33"},"PeriodicalIF":1.0,"publicationDate":"2022-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88841589","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Tianxing Li, Jin Huang, Erik Risinger, Deepak Ganesan
While multi-modal deep learning is useful in distributed sensing tasks like human tracking, activity recognition, and audio and video analysis, deploying state-of-the-art multi-modal models in a wirelessly networked sensor system poses unique challenges. The data sizes for different modalities can be highly asymmetric (e.g., video vs. audio), and these differences can lead to significant delays between streams in the presence of wireless dynamics. Therefore, a slow stream can significantly slow down a multimodal inference system in the cloud, leading to either increased latency (when blocked by the slow stream) or degradation in inference accuracy (if inference proceeds without waiting).
{"title":"Low-Latency Speculative Inference on Distributed Multi-Modal Data Streams","authors":"Tianxing Li, Jin Huang, Erik Risinger, Deepak Ganesan","doi":"10.1145/3568113.3568121","DOIUrl":"https://doi.org/10.1145/3568113.3568121","url":null,"abstract":"While multi-modal deep learning is useful in distributed sensing tasks like human tracking, activity recognition, and audio and video analysis, deploying state-of-the-art multi-modal models in a wirelessly networked sensor system poses unique challenges. The data sizes for different modalities can be highly asymmetric (e.g., video vs. audio), and these differences can lead to significant delays between streams in the presence of wireless dynamics. Therefore, a slow stream can significantly slow down a multimodal inference system in the cloud, leading to either increased latency (when blocked by the slow stream) or degradation in inference accuracy (if inference proceeds without waiting).","PeriodicalId":29918,"journal":{"name":"GetMobile-Mobile Computing & Communications Review","volume":"43 1","pages":"23 - 26"},"PeriodicalIF":1.0,"publicationDate":"2022-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81103547","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：