Improving Object Localization with Fitness NMS and Bounded IoU Loss

Select | Print


Tychsen-Smith, Lachlan; Petersson, Lars


2018-06-18


Conference Material


Conference on Computer Vision and Pattern Recognition, Salt Lake City, USA, 18–22 June 2018


6877-6885


We demonstrate that many detection methods are designed to identify only a sufficently accurate bounding box, rather than the best available one. To address this issue we propose a simple and fast modification to the existing methods called Fitness NMS. This method is tested with the DeNet model and obtains a significantly improved MAP at greater localization accuracies without a loss in evaluation rate, and can be used in conjunction with Soft NMS for additional improvements. Next we derive a novel bounding box regression loss based on a set of IoU upper bounds that better matches the goal of IoU maximization while still providing good convergence properties. Following these novelties we investigate RoI clustering schemes for improving evaluation rates for the DeNet wide model variants and provide an analysis of localization performance at various input image dimensions. We obtain a MAP of 33.6%@79Hz and 41.8%@5Hz for MSCOCO and a Titan X (Maxwell).


Institute of Electrical and Electronics Engineers


CVPR2018


computer vision, object detection, neural networks, deep learning


Computer Vision


https://doi.org/10.1109/CVPR.2018.00719


EP183692


Conference Paper - Refereed


English


Tychsen-Smith, Lachlan; Petersson, Lars. Improving Object Localization with Fitness NMS and Bounded IoU Loss. In: Conference on Computer Vision and Pattern Recognition; 18–22 June 2018; Salt Lake City, USA. CVPR2018: Institute of Electrical and Electronics Engineers; 2018. 6877-6885. https://doi.org/10.1109/CVPR.2018.00719



Loading citation data...

Citation counts
(Requires subscription to view)