Closed seekingdeep closed 3 years ago
@seekingdeep We conduct no further tuning of hyper-parameters besides the work of the paper, so the results should be the same as the online-version paper. The FPS is estimated at single-scale inputs while the mIoU score is based on multi-scale testing, following [1].
[1] C. Yu, J. Wang, C. Peng, C. Gao, G. Yu, and N. Sang, “BiSeNet: Bilateral segmentation network for real-time semantic segmentation,” in European Conference on Computer Vision, 2018, pp. 334–349.
@BebDong Hi there, what are the results of EPRNet parameters, flops, miou, fps at what scale size?