DeepKey: Towards End-to-End Physical Key Replication From a Single Photograph

Rory Smith; Tilo Burghardt

doi:10.1007/978-3-030-12939-2_34

DeepKey: Towards End-to-End Physical Key Replication From a Single Photograph

Rory Smith^*, Tilo Burghardt

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference Contribution (Conference Proceeding)

80 Downloads (Pure)

Abstract

This paper describes DeepKey, an end-to-end deep neural architecture capable of taking a digital RGB image of an ‘everyday’ scene containing a pin tumbler key (e.g. lying on a table or carpet) and fully automatically inferring a printable 3D key model. We report on the key detection performance and describe how candidates can be transformed into physical prints. We show an example opening a real-world lock. Our system is described in detail, providing a breakdown of all components including key detection, pose normalisation, bitting segmentation and 3D model inference. We provide an in-depth evaluation and conclude by reﬂecting on limitations, applications, potential security risks and societal impact. We contribute the DeepKey Datasets of 5,300+ images covering a few test keys with bounding boxes, pose and unaligned mask data.

Original language	English
Title of host publication	Pattern Recognition
Subtitle of host publication	40th German Conference, GCPR 2018, Stuttgart, Germany, October 9-12, 2018, Proceedings
Editors	Thomas Brox, Andrés Bruhn, Mario Fritz
Publisher	Springer, Cham
Pages	487-502
Number of pages	16
ISBN (Electronic)	9783030129392
ISBN (Print)	9783030129385
DOIs	https://doi.org/10.1007/978-3-030-12939-2_34
Publication status	Published - 14 Feb 2019
Event	40th German Conference on Pattern Recognition, GCPR 2018 - Stuttgart, Germany Duration: 9 Oct 2018 → 12 Oct 2018

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11269 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	40th German Conference on Pattern Recognition, GCPR 2018
Country/Territory	Germany
City	Stuttgart
Period	9/10/18 → 12/10/18

Access to Document

10.1007/978-3-030-12939-2_34

Full-text PDF (accepted author manuscript)
This is the author accepted manuscript (AAM). The final published version (version of record) is available online via Springer Nature at https://link.springer.com/chapter/10.1007%2F978-3-030-12939-2_34. Please refer to any applicable terms of use of the publisher.
Accepted author manuscript, 5.85 MB

Handle.net

Persistent link

Cite this

Smith, R., & Burghardt, T. (2019). DeepKey: Towards End-to-End Physical Key Replication From a Single Photograph. In T. Brox, A. Bruhn, & M. Fritz (Eds.), Pattern Recognition: 40th German Conference, GCPR 2018, Stuttgart, Germany, October 9-12, 2018, Proceedings (pp. 487-502). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11269 LNCS). Springer, Cham. https://doi.org/10.1007/978-3-030-12939-2_34

Smith, Rory ; Burghardt, Tilo. / DeepKey : Towards End-to-End Physical Key Replication From a Single Photograph. Pattern Recognition: 40th German Conference, GCPR 2018, Stuttgart, Germany, October 9-12, 2018, Proceedings. editor / Thomas Brox ; Andrés Bruhn ; Mario Fritz. Springer, Cham, 2019. pp. 487-502 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{39c6524331064337bfbcb1a43e64270e,

title = "DeepKey: Towards End-to-End Physical Key Replication From a Single Photograph",

abstract = "This paper describes DeepKey, an end-to-end deep neural architecture capable of taking a digital RGB image of an {\textquoteleft}everyday{\textquoteright} scene containing a pin tumbler key (e.g. lying on a table or carpet) and fully automatically inferring a printable 3D key model. We report on the key detection performance and describe how candidates can be transformed into physical prints. We show an example opening a real-world lock. Our system is described in detail, providing a breakdown of all components including key detection, pose normalisation, bitting segmentation and 3D model inference. We provide an in-depth evaluation and conclude by reﬂecting on limitations, applications, potential security risks and societal impact. We contribute the DeepKey Datasets of 5,300+ images covering a few test keys with bounding boxes, pose and unaligned mask data.",

author = "Rory Smith and Tilo Burghardt",

year = "2019",

month = feb,

day = "14",

doi = "10.1007/978-3-030-12939-2_34",

language = "English",

isbn = "9783030129385",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer, Cham",

pages = "487--502",

editor = "Thomas Brox and Andr{\'e}s Bruhn and Mario Fritz",

booktitle = "Pattern Recognition",

address = "Switzerland",

note = "40th German Conference on Pattern Recognition, GCPR 2018 ; Conference date: 09-10-2018 Through 12-10-2018",

}

Smith, R & Burghardt, T 2019, DeepKey: Towards End-to-End Physical Key Replication From a Single Photograph. in T Brox, A Bruhn & M Fritz (eds), Pattern Recognition: 40th German Conference, GCPR 2018, Stuttgart, Germany, October 9-12, 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11269 LNCS, Springer, Cham, pp. 487-502, 40th German Conference on Pattern Recognition, GCPR 2018, Stuttgart, Germany, 9/10/18. https://doi.org/10.1007/978-3-030-12939-2_34

DeepKey: Towards End-to-End Physical Key Replication From a Single Photograph. / Smith, Rory; Burghardt, Tilo.
Pattern Recognition: 40th German Conference, GCPR 2018, Stuttgart, Germany, October 9-12, 2018, Proceedings. ed. / Thomas Brox; Andrés Bruhn; Mario Fritz. Springer, Cham, 2019. p. 487-502 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11269 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference Contribution (Conference Proceeding)

TY - GEN

T1 - DeepKey

T2 - 40th German Conference on Pattern Recognition, GCPR 2018

AU - Smith, Rory

AU - Burghardt, Tilo

PY - 2019/2/14

Y1 - 2019/2/14

N2 - This paper describes DeepKey, an end-to-end deep neural architecture capable of taking a digital RGB image of an ‘everyday’ scene containing a pin tumbler key (e.g. lying on a table or carpet) and fully automatically inferring a printable 3D key model. We report on the key detection performance and describe how candidates can be transformed into physical prints. We show an example opening a real-world lock. Our system is described in detail, providing a breakdown of all components including key detection, pose normalisation, bitting segmentation and 3D model inference. We provide an in-depth evaluation and conclude by reﬂecting on limitations, applications, potential security risks and societal impact. We contribute the DeepKey Datasets of 5,300+ images covering a few test keys with bounding boxes, pose and unaligned mask data.

AB - This paper describes DeepKey, an end-to-end deep neural architecture capable of taking a digital RGB image of an ‘everyday’ scene containing a pin tumbler key (e.g. lying on a table or carpet) and fully automatically inferring a printable 3D key model. We report on the key detection performance and describe how candidates can be transformed into physical prints. We show an example opening a real-world lock. Our system is described in detail, providing a breakdown of all components including key detection, pose normalisation, bitting segmentation and 3D model inference. We provide an in-depth evaluation and conclude by reﬂecting on limitations, applications, potential security risks and societal impact. We contribute the DeepKey Datasets of 5,300+ images covering a few test keys with bounding boxes, pose and unaligned mask data.

UR - http://www.scopus.com/inward/record.url?scp=85063512210&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-12939-2_34

DO - 10.1007/978-3-030-12939-2_34

M3 - Conference Contribution (Conference Proceeding)

AN - SCOPUS:85063512210

SN - 9783030129385

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 487

EP - 502

BT - Pattern Recognition

A2 - Brox, Thomas

A2 - Bruhn, Andrés

A2 - Fritz, Mario

PB - Springer, Cham

Y2 - 9 October 2018 through 12 October 2018

ER -

Smith R, Burghardt T. DeepKey: Towards End-to-End Physical Key Replication From a Single Photograph. In Brox T, Bruhn A, Fritz M, editors, Pattern Recognition: 40th German Conference, GCPR 2018, Stuttgart, Germany, October 9-12, 2018, Proceedings. Springer, Cham. 2019. p. 487-502. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-12939-2_34

DeepKey: Towards End-to-End Physical Key Replication From a Single Photograph

Abstract

Publication series

Conference

Access to Document

Handle.net

Other files and links

Fingerprint

Cite this