@knwl_cmp_lab

Research Group at the University of Minnesota, advancing the frontier of knowledge at the intersection of Machine Learning and Spatial Data

knowledge-computing.github.io/
Joined February 2024
  • Tweets35
  • Following3
  • Followers16
8Media
Welcome to our research group! We are currently focusing on building machine learning frameworks based on spatial data to address challenges in natural language processing, computer vision, and spatiotemporal data prediction & forecasting.
2
217
Knowledge Computing Lab retweeted
ICDAR MapText Competition@ICDAR_MapText6 Mar 2025
2/2 The winning methods will be featured in the ICDAR Proceedings Paper, and winners will have the opportunity to receive certificates on stage at the ICDAR conference in Wuhan!🥳
3
2
217
Knowledge Computing Lab retweeted
ICDAR MapText Competition@ICDAR_MapText6 Mar 2025
1/2 ✨ICDAR25 MapText Comp has released the test dataset✨, with submission deadline at April 1, 2025. 👉 rrc.cvc.uab.es/?ch=32&com=in…
4
1
2
326
Dr. @yaoyichiang, @leeje_0413, and @jina_kimmm recently visited @EwhaWomansUniv to present their ongoing research. We sincerely thank Prof. Jaehyung Sim for hosting the event and facilitating the exchange of valuable insights between the two labs’ current research endeavors.
71
Knowledge Computing Lab retweeted
ICDAR MapText Competition@ICDAR_MapText17 Jan 2025
ICDAR MapText Competition is back for 2025 with new datasets and much more! 🥳 Winners will be awarded and announced at the ICDAR conference.📷Website: rrc.cvc.uab.es/?ch=32
1
87
Thrilled to share that Rhett Olson, an undergraduate researcher in our lab, has been selected for Honorable Mention for the 2024-2025 Outstanding Undergraduate Researcher Award from the @CRAtweets! Congratulations, Rhett! 🎉👏 #UMNProud @UMNCompsci
25
🚨 Join @cugeography as a Postdoctoral Associate to develop advanced methods for extracting historical geographic information from digital archived maps (1884-2006). 🗓️ Apply by Dec 16, 2024; open until filled. Details & apply 👉 bit.ly/Job60562
43
🚨Exciting Announcement!🚨 @jina_kimmm and @Yijun_Lin0 will co-instructing 'CSCI5523: Introduction to Data Mining' this Spring 2025!📊✨ Join the course to learn essential techniques for discovering patterns and insights from data from our two great instructors.💻🔍
1
47
It was an exciting week at SIGSPATIAL'24 conference in Atlanta, Georgia! Our lab members had the opportunity to present their latest work and connect with other researchers in spatial computing. Thank you to everyone who engaged with us during the conference!
1
1
112
🏙️We are excited to share 'CrossBag: A Bag of Tricks for Cross-City Mobility Prediction' at @SIGSPATIAL_GIS HuMob'24 🗓️ Date: Oct. 29, TBD Congratulations to JangHyeon Lee and @yaoyichiang 🎉
1
75
What’s in my CrossBag? A set of simple yet effective techniques (before/during/after training) to boost cross-city human mobility (HuMob) prediction, placing in the top 10 out of 100+ teams in the HuMob Challenge 2024. #SIGSPATIAL2024 #UMNProud @UMNComputerSci
86
🌳 We are thrilled to share our research on 'MiTREE: Multi-input Transformer Ecoregion Encoder for Species Distribution Modelling' at @SIGSPATIAL_GIS GeoAI'24 🗓️ Date: Oct. 29, 2:15-2:30 📍 Location: Salon 4 Congratulations to Thresa Chen and @yaoyichiang 🎉
1
61
MiTREE is a multi-input vision-transformer model that integrates ecological classifications and location data to improve bird species predictions. #SIGSPATIAL2024 #UMNProud @UMNComputerSci
65
📷 We are excited to present our research titled 'Enabling Semantic-Rich Location Search on Street View Imagery Using Multilingual POI Data' at @SIGSPATIAL_GIS GeoSearch '24 🗓️ Oct. 29, 10:50-11:10 Congrats to Leeje Jang and @yaoyichiang 🎉
1
48
We introduce MulMapper, a system that enhances street view imagery by extracting multilingual text, detecting text alignment, and matching POI attributes to text in the image, significantly improving the performance of location-based searches.
1
38
#SIGSPATIAL2024 #UMNProud @UMNComputerSci
39
🗺️ We are excited to share our research on 'Automatic Search of Multiword Place Names on Historical Maps' at @SIGSPATIAL_GIS GeoSearch'24 🗓️Date: Oct. 29, 9:00-9:20 📍Location: Conference E Congratulations Rhett Olson, @jina_kimmm, and @yaoyichiang 🎉
1
57
We present a method to efficiently search multiword place names on historical maps by linking text labels using a minimum spanning tree. It connects words based on spatial proximity, height, angle, and capitalization to retrieve specified place names.
1
56
#SIGSPATIAL2024 #UMNProud @UMNComputerSci
58
🚀 Excited to share our research on Context-Aware Trajectory Anomaly Detection at @SIGSPATIAL_GIS GeoAnomalies’24! 🗓️ Date: Oct. 29 11:00-11:20 📍 Location: Conference Room 2 Congrats to Haoji Hu, @jina_kimmm, Jinwei Zhou, @sofia_kirsanova, Janghyeon Lee, and @yaoyichiang 🎉
1
3
3
361
Our paper presents a context-aware approach that integrates agent IDs and Points of Interest (POI) data, enabling more accurate agent-level anomaly detection in trajectories. #SIGSPATIAL2024 #UMNProud
1
3
179
Load more