Sketch recognition describes the process by which a computer, or artificial intelligence can interpret hand-drawn sketches created by a human being, or other machine.[1] Sketch recognition is a key frontier in the field of artificial intelligence and human-computer interaction, similar to natural language processing or conversational artificial intelligence[2][3]

Uses and Applications

edit

Research in sketch recognition lies at the crossroads of artificial intelligence and human–computer interaction. Recognition algorithms usually are gesture-based, appearance-based, geometry-based, or a combination thereof.

Advances in the field of sketch recognition would have significant application in the field of forensic science, in which sketches are often used to identify suspects associated with a crime.[4][5]

In 2023, two developers used OpenAI's DallE-2 image generation platform to create a forensic sketch program. The program's results were described as "hyper-realistic" and purported the potential of exponentially decreasing the creation time of a forensic sketch, while increasing accuracy.[6]

Sketch recognition technology has also been linked to applications in the fields of architecture, videogame production, animation, construction, and academia, among others.[7][8][9][10]

See also

edit

References

edit
  1. ^ Hammond, T. and Davis, R. (2005), "LADDER, a sketching language for user interface developers", Computers & Graphics, 2005, 29(4), pp. 518-532.
  2. ^ Hammond, T., Logsdon, D., Peschel, J., Johnston, J., Taele, P., Wolin, A., and Paulson, B. A sketch recognition interface that recognizes hundreds of shapes in course-of-action diagrams. In Proceedings of the 28th of the international conference extended abstracts on Human factors in computing systems (CHI EA '10), 2010, pp. 4213-4218.
  3. ^ Jorge, J. and Samavati, F. (2011), Sketch-Based Interfaces and Modeling, Springer
  4. ^ IJRASET. "Forensic Face Sketch Construction and Recognition". www.ijraset.com. Retrieved 2024-01-10.
  5. ^ Lei, Haopeng; Chen, Simin; Wang, Mingwen; He, Xiangjian; Jia, Wenjing; Li, Sibo (2021-05-25). "A New Algorithm for Sketch-Based Fashion Image Retrieval Based on Cross-Domain Transformation". Wireless Communications and Mobile Computing. 2021: e5577735. doi:10.1155/2021/5577735. hdl:10453/149881. ISSN 1530-8669.
  6. ^ Xiang, Chloe (2023-02-07). "Developers Created AI to Generate Police Sketches. Experts Are Horrified". Vice. Retrieved 2024-01-10.
  7. ^ "Papers with Code - Sketch Recognition". paperswithcode.com. Retrieved 2024-01-10.
  8. ^ "Sketch Recognition". Microsoft Research. Retrieved 2024-01-10.
  9. ^ Zhang, Lei (2021-08-21). "Hand-drawn sketch recognition with a double-channel convolutional neural network". EURASIP Journal on Advances in Signal Processing. 2021 (1): 73. Bibcode:2021EJASP2021...73Z. doi:10.1186/s13634-021-00752-4. ISSN 1687-6180.
  10. ^ Korkut, Elif Hilal; Surer, Elif (2021). "Sketch Recognition for Interactive Game Experiences Using Neural Networks". In Baalsrud Hauge, Jannicke; C. S. Cardoso, Jorge; Roque, Licínio; Gonzalez-Calero, Pedro A. (eds.). Entertainment Computing – ICEC 2021. Lecture Notes in Computer Science. Cham: Springer International Publishing. pp. 393–401. doi:10.1007/978-3-030-89394-1_31. ISBN 978-3-030-89394-1. S2CID 240415939.
edit