no code implementations • 16 Oct 2023 • Pierre Le Jeune
We demonstrate this with an in-depth analysis of existing FSOD methods on aerial images and observed a large performance gap compared to natural images.
no code implementations • 17 Jul 2023 • Pierre Le Jeune, Anissa Mokraoui
SIoU improves small object detection in the non-few-shot regime, but this setting is unrealistic in the industry as annotated detection datasets are often too expensive to acquire.
1 code implementation • 25 Oct 2022 • Pierre Le Jeune, Anissa Mokraoui
Finally, we develop a novel multiscale alignment method, Cross-Scales Query-Support Alignment (XQSA) for FSOD, to improve the detection of small objects.
no code implementations • 6 Jan 2022 • Pierre Le Jeune, Anissa Mokraoui
In particular, for attention-based FSOD methods, it is laborious to compare the impact of the different attention mechanisms on performance.
1 code implementation • 27 Sep 2021 • Pierre Le Jeune, Mustapha Lebbah, Anissa Mokraoui, Hanene Azzag
This training strategy encourages the network to adapt to new classes as it would at test time.