Revision as of 04:15, 26 September 2022 edit Citation bot (talk \| contribs) Bots 5,868,599 edits Alter: template type. Add: class. \| Use this bot. Report bugs. \| Suggested by Headbomb \| #UCB_toolbar ← Previous edit		Revision as of 14:10, 3 January 2023 edit undo Citation bot (talk \| contribs) Bots 5,868,599 edits Alter: title, template type. Add: s2cid, isbn, arxiv, doi, pages, chapter. Removed parameters. Some additions/deletions were parameter name changes. \| Use this bot. Report bugs. \| Suggested by SemperIocundus \| #UCB_webform 888/2500 Next edit →
Line 43: === Add-on techniques === Instead of modifying existing methods, some add-on techniques are there, which can be directly placed on top of existing approaches to detect smaller objects. One such technique is Slicing Aided Hyper Inference(SAHI).<ref>{{cite ~~arXiv~~book \|last1=Akyon \|first1=Fatih Cagatay \|last2=Altinuc \|first2=Sinan Onur \|last3=Temizel \|first3=Alptekin \|date=2022-07-12 \|title=2022 IEEE International Conference on Image Processing (ICIP) \|chapter=Slicing Aided Hyper Inference and Fine-~~tuning~~Tuning for Small Object Detection \|~~class~~pages=cs966–970 \|doi=10.1109/ICIP46576.2022.CV9897990 \|~~eprint~~arxiv=2202.06934\|isbn=978-1-6654-9620-9 \|s2cid=246823962 }}</ref> The image is sliced into different-sized multiple overlapping patches. [[Hyperparameter (machine learning)\|Hyper-parameters]] define their dimensions. Then patches are resized, while maintaining the aspect ratio during fine-tuning. These patches are then provided for training the model. === Well-Optimised techniques for small object detection ===

Small object detection: Difference between revisions