Mobile computing | Ramesh Govindan

Our work has considered a range of issues in mobile computing, from processing visual information on mobile devices, to reliably programming them, and leverage cloud and edge resources for applications.

Multimedia
SplatPose: On-Device Outdoor AR Pose Estimation Using Gaussian Splatting

Weiwu Pang, Rajrup Ghosh, Jiawei Yang, and 4 more authors

In Proceedings of the 33rd ACM International Conference on Multimedia, 2025

Abs Bib Link

Outdoor AR applications on mobile devices need accurate estimates for the pose of the device. In this paper, we develop SplatPose, a novel pose estimation technique that uses a data-driven 3D modeling technique called Gaussian Splatting. SplatPose uses a trained Gaussian Splatting model to render an image at an estimated device location, then matches features with the camera image to estimate pose. % Because this matching can be fast, SplatPose can, in theory, estimate pose entirely on a mobile device, while existing approaches cannot. To this end, SplatPose trains Gaussian Splatting models to be robust to appearance changes, thereby improving accuracy. It also incorporates a novel fast renderer to improve rendering speed. Using an AR pose estimation benchmark dataset, we show that SplatPose outperforms the state-of-the-art in terms of accuracy, and is up to an order of magnitude faster on a mobile device.
@inproceedings{pang2025, author = {Pang, Weiwu and Ghosh, Rajrup and Yang, Jiawei and Wei, Ziyu and Leong, Branden and Wang, Yue and Govindan, Ramesh}, title = {SplatPose: On-Device Outdoor AR Pose Estimation Using Gaussian Splatting}, booktitle = {Proceedings of the 33rd ACM International Conference on Multimedia}, year = {2025}, series = {MM '25}, pages = {11899–11908}, address = {New York, NY, USA}, publisher = {Association for Computing Machinery}, isbn = {9798400720352}, url = {https://doi.org/10.1145/3746027.3755709}, doi = {10.1145/3746027.3755709}, numpages = {10}, keywords = {augmented reality, gaussian splatting, outdoor pose estimation}, topics = {mobile}, location = {Dublin, Ireland} }

IoTDI

Rim: Offloading Inference to the Edge

Yitao Hu, Weiwu Pang, Xiaochen Liu, and 4 more authors

In Proceedings of the 6th ACM/IEEE Conference on Internet of Things Design and Implementation, 2021, 2021

Bib Link

@inproceedings{hu.y.pang.w.ea:rim,
  title = {Rim: Offloading Inference to the Edge},
  author = {Hu, Yitao and Pang, Weiwu and Liu, Xiaochen and Ghosh, Rajrup and Ko, Bongjun and Lee, Wei-Han and Govindan, Ramesh},
  url = {https://dl.acm.org/doi/abs/10.1145/3450268.3453521},
  year = {2021},
  date = {2021-05-18},
  booktitle = {Proceedings of the 6th ACM/IEEE Conference on Internet of
  		  Things Design and Implementation, 2021},
  pubstate = {published},
  topics = {sensing,mobile},
  tppubtype = {inproceedings}
}

SoCC

Scrooge: A Cost-Effective Deep Learning Inference System

Yitao Hu, Rajrup Ghosh, and Ramesh Govindan

In SoCC ’21: ACM Symposium on Cloud Computing, Seattle, WA, USA, November 1 - 4, 2021, 2021

Bib Link

@inproceedings{hu.y.ghosh.r.ea:scrooge,
  author = {Hu, Yitao and Ghosh, Rajrup and Govindan, Ramesh},
  editor = {Curino, Carlo and Koutrika, Georgia and Netravali, Ravi},
  title = {Scrooge: {A} Cost-Effective Deep Learning Inference
  		  System},
  booktitle = {SoCC '21: {ACM} Symposium on Cloud Computing, Seattle, WA,
  		  USA, November 1 - 4, 2021},
  pages = {624--638},
  publisher = {{ACM}},
  year = {2021},
  url = {https://doi.org/10.1145/3472883.3486993},
  doi = {10.1145/3472883.3486993},
  timestamp = {Sun, 31 Oct 2021 10:05:01 +0100},
  topics = {sensing,mobile},
  biburl = {https://dblp.org/rec/conf/cloud/HuGG21.bib}
}

Middleware

Olympian: Scheduling GPU Usage in a Deep Neural Network Model Serving System

Yitao Hu, Swati Rallapalli, Bongjun Ko, and 1 more author

In Proceedings of the 19th International Middleware Conference, 2018

Bib Link

@inproceedings{hu.y.rallapalli.s.ea:olympian,
  title = {Olympian: Scheduling GPU Usage in a Deep Neural Network
  		  Model Serving System},
  author = {Hu, Yitao and Rallapalli, Swati and Ko, Bongjun and Govindan, Ramesh},
  url = {http://doi.acm.org/10.1145/3274808.3274813},
  doi = {10.1145/3274808.3274813},
  isbn = {978-1-4503-5702-9},
  year = {2018},
  date = {2018-12-01},
  booktitle = {Proceedings of the 19th International Middleware
  		  Conference},
  pages = {53--65},
  publisher = {ACM},
  address = {Rennes, France},
  series = {Middleware '18},
  keywords = {Deep Neural Network, GPU Scheduling, Multiple Concurrent
  		  Requests, Predictable Latency, Scheduling Algorithms,
  		  Serving System},
  pubstate = {published},
  topics = {mobile},
  tppubtype = {inproceedings}
}

MobiSys

Gnome: A Practical Approach to NLOS Mitigation for GPS Positioning in Smartphones

Xiaochen Liu, Suman Nath, and Ramesh Govindan

In Proceedings of the 16th Annual International Conference on Mobile Systems, Applications, and Services (Mobisys), 2018

Bib Link

@inproceedings{liu.x.nath.s.ea:gnome,
  title = {Gnome: A Practical Approach to NLOS Mitigation for GPS
  		  Positioning in Smartphones},
  author = {Liu, Xiaochen and Nath, Suman and Govindan, Ramesh},
  url = {http://doi.acm.org/10.1145/3210240.3210343},
  doi = {10.1145/3210240.3210343},
  isbn = {978-1-4503-5720-3},
  year = {2018},
  date = {2018-01-01},
  booktitle = {Proceedings of the 16th Annual International Conference on
  		  Mobile Systems, Applications, and Services (Mobisys)},
  pages = {163--177},
  publisher = {ACM},
  address = {Munich, Germany},
  series = {MobiSys '18},
  topics = {mobile},
  keywords = {GPS, Localization, Mobile Computing, NLOS Mitigation},
  pubstate = {published},
  tppubtype = {inproceedings}
}

Ubicomp

ALPS: Accurate Landmark Positioning at City Scales

Yitao Hu, Xiaochen Liu, Suman Nath, and 1 more author

In the 2016 ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp 2016) , se 2016

Bib

@inproceedings{hu.y.liu.x.ea:alps--accurate-landmark-positioning-at-city-scales,
  author = {Hu, Yitao and Liu, Xiaochen and Nath, Suman and Govindan, Ramesh},
  title = { {ALPS: Accurate Landmark Positioning at City Scales} },
  booktitle = { the 2016 ACM International Joint Conference on Pervasive
  		  and Ubiquitous Computing (UbiComp 2016) },
  year = {2016},
  month =  se,
  location = { Heidelberg, Germany },
  topics = {mobile},
  		  }
}

MobiSys

Efficient Privilege De-Escalation for Ad Libraries in Mobile Apps

Bin Liu, Bin Liu, Hongxia Jin, and 1 more author

In The 13th Annual International Conference on Mobile Systems, Applications, and Services (MobiSys 2015) , ma 2015

Bib

@inproceedings{liu.b.liu.b.ea:efficient-privilege-de-escalation-for-ad-libraries-in-mobile-apps,
  author = {Liu, Bin and Liu, Bin and Jin, Hongxia and Govindan, Ramesh},
  title = { {Efficient Privilege De-Escalation for Ad Libraries in
  		  Mobile Apps} },
  booktitle = { The 13th Annual International Conference on Mobile
  		  Systems, Applications, and Services (MobiSys 2015) },
  month =  ma,
  location = { Florence, Italy },
  topics = {mobile},
  year = {2015},
  projects = { }
}

MobiSys

FlexiWeb: Network-Aware Compaction for Accelerating Mobile Web Transfers

S. Singh, H. Madhyastha, S. Krishnamurthy, and 1 more author

In Proc. ACM MobiCom, ma 2015

Bib

@inproceedings{singh.s.madhyastha.h.ea:flexiweb--network-aware-compaction-for-accelerating-mobile-web-transfers,
  author = {Singh, S. and Madhyastha, H. and Krishnamurthy, S. and Govindan, R.},
  title = {{FlexiWeb: Network-Aware Compaction for Accelerating
  		  Mobile Web Transfers}},
  booktitle = {Proc. ACM MobiCom},
  year = {2015},
  topics = {mobile},
  address = {Paris, France}
}