From c810848c514c4bf9af870b1a388c3ddeb914e7ce Mon Sep 17 00:00:00 2001 From: zhoujun <572459439@qq.com> Date: Fri, 3 Jul 2020 09:39:10 +0800 Subject: [PATCH] =?UTF-8?q?=E6=B7=BB=E5=8A=A0=E6=B3=A8=E9=87=8A?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- torchocr/deprecated/FeaturePyramidNetwork.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/torchocr/deprecated/FeaturePyramidNetwork.py b/torchocr/deprecated/FeaturePyramidNetwork.py index 9dda213..4b5e8b5 100644 --- a/torchocr/deprecated/FeaturePyramidNetwork.py +++ b/torchocr/deprecated/FeaturePyramidNetwork.py @@ -3,7 +3,9 @@ import torch.nn.functional as F from torch import nn, Tensor - +""" +out_channels=96时,和现有的fpn相比,这个fpn能提升0.01%的精度,但是模型尺寸会大500k +""" class FeaturePyramidNetwork(nn.Module): def __init__(self, in_channels, out_channels=256):