SphOR: A Representation Learning Perspective on Open-set Recognition for Identifying Unknown Classes in Deep Learning Models.
Nadarasar Bahavan, Sachith Seneviratne, Saman Halgamuge, AI, Optimization, Pattern Recognition Group, The University of Melbourne [email protected], {sachith.seneviratne,saman.halgamuge}@unimelb.edu.au · 2025
Read SphOR: A Representation Learning Perspective on Open-set Recognition for Identifying Unknown Classes in Deep Learning Models. by Nadarasar Bahavan, Sachith Seneviratne, Saman Halgamuge, AI, Optimization, Pattern Recognition Group, The University of Melbourne [email protected], {sachith.seneviratne,saman.halgamuge}@unimelb.edu.au (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-08049.txt.
Description
Nadarasar Bahavan and co-authors teach deep nets to say 'new kind' when a shot fits no class they know. Closed nets force each shot into a known bin, a risk in safe-first fields. Their 2025 way, SphOR, shapes the eye's space first and fits the judge last. Class marks are set square to each other so kinds stand far apart. Shots are then held on a ball skin as bell blends.
Shake tricks ride the train to blunt false faith. Open-set craft past this pair-fit the eye and judge as one, so the eye bent odd for new kinds. Stock sight builds, built for broad tasks, bent no less. A space shaped for the open task can flag new kinds clean, they hold.
Catalog record
| CREATOR | Nadarasar Bahavan, Sachith Seneviratne, Saman Halgamuge, AI, Optimization, Pattern Recognition Group, The University of Melbourne [email protected], {sachith.seneviratne,saman.halgamuge}@unimelb.edu.au |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 40 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-08049