Show simple item record

dc.contributor.authorMohamed, Abduallah
dc.contributor.authorZhu, Deyao
dc.contributor.authorVu, Warren
dc.contributor.authorElhoseiny, Mohamed
dc.contributor.authorClaudel, Christian
dc.date.accessioned2022-12-05T12:09:05Z
dc.date.available2022-05-16T08:59:30Z
dc.date.available2022-12-05T12:09:05Z
dc.date.issued2022-10-23
dc.identifier.citationMohamed, A., Zhu, D., Vu, W., Elhoseiny, M., & Claudel, C. (2022). Social-Implicit: Rethinking Trajectory Prediction Evaluation and The Effectiveness of Implicit Maximum Likelihood Estimation. Computer Vision – ECCV 2022, 463–479. https://doi.org/10.1007/978-3-031-20047-2_27
dc.identifier.isbn9783031200465
dc.identifier.issn1611-3349
dc.identifier.issn0302-9743
dc.identifier.doi10.1007/978-3-031-20047-2_27
dc.identifier.urihttp://hdl.handle.net/10754/677949
dc.description.abstractBest-of-N (BoN) Average Displacement Error (ADE)/ Final Displacement Error (FDE) is the most used metric for evaluating trajectory prediction models. Yet, the BoN does not quantify the whole generated samples, resulting in an incomplete view of the model’s prediction quality and performance. We propose a new metric, Average Mahalanobis Distance (AMD) to tackle this issue. AMD is a metric that quantifies how close the whole generated samples are to the ground truth. We also introduce the Average Maximum Eigenvalue (AMV) metric that quantifies the overall spread of the predictions. Our metrics are validated empirically by showing that the ADE/FDE is not sensitive to distribution shifts, giving a biased sense of accuracy, unlike the AMD/AMV metrics. We introduce the usage of Implicit Maximum Likelihood Estimation (IMLE) as a replacement for traditional generative models to train our model, Social-Implicit. IMLE training mechanism aligns with AMD/AMV objective of predicting trajectories that are close to the ground truth with a tight spread. Social-Implicit is a memory efficient deep model with only 5.8K parameters that runs in real time of 580 Hz and achieves competitive results.
dc.publisherSpringer Nature Switzerland
dc.relation.urlhttps://link.springer.com/10.1007/978-3-031-20047-2_27
dc.rightsThis is an accepted manuscript version of a paper before final publisher editing and formatting. Archived with thanks to Springer Nature Switzerland.
dc.titleSocial-Implicit: Rethinking Trajectory Prediction Evaluation and The Effectiveness of Implicit Maximum Likelihood Estimation
dc.typeConference Paper
dc.contributor.departmentComputer, Electrical and Mathematical Science and Engineering (CEMSE) Division
dc.contributor.departmentComputer Science Program
dc.contributor.departmentVisual Computing Center (VCC)
dc.rights.embargodate2023-10-23
dc.conference.date2022-10-23 to 2022-10-27
dc.conference.name17th European Conference on Computer Vision, ECCV 2022
dc.conference.locationTel Aviv, ISR
dc.eprint.versionPost-print
dc.contributor.institutionThe University of Texas, Austin, USA
dc.identifier.volume13682 LNCS
dc.identifier.pages463-479
dc.identifier.arxivid2203.03057
kaust.personZhu, Deyao
kaust.personElhoseiny, Mohamed
dc.relation.issupplementedbygithub:abduallahmohamed/Social-Implicit
dc.identifier.eid2-s2.0-85142694428
refterms.dateFOA2022-05-16T10:10:37Z
display.relations<b>Is Supplemented By:</b><br/> <ul><li><i>[Software]</i> <br/> Title: abduallahmohamed/Social-Implicit: Code for: "Social-Implicit: Rethinking Trajectory Prediction Evaluation and The Effectiveness of Implicit Maximum Likelihood Estimation" Accepted @ ECCV2022. Publication Date: 2021-11-22. github: <a href="https://github.com/abduallahmohamed/Social-Implicit" >abduallahmohamed/Social-Implicit</a> Handle: <a href="http://hdl.handle.net/10754/686453" >10754/686453</a></a></li></ul>


Files in this item

Thumbnail
Name:
2203.03057 (1).pdf
Size:
3.374Mb
Format:
PDF
Description:
Accepted Manuscript
Embargo End Date:
2023-10-23

This item appears in the following Collection(s)

Show simple item record

VersionItemEditorDateSummary

*Selected version