From 32a87b44acd6f9a0f76ed2caf401320f239fcc77 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E6=B2=99=E9=9F=AC=E4=BC=9F?= <11112877@vivo.xyz>
Date: Tue, 4 Aug 2020 20:38:23 +0800
Subject: [PATCH] =?UTF-8?q?=E5=8F=AC=E5=9B=9E=E9=97=AE=E9=A2=98=E8=A1=A5?=
 =?UTF-8?q?=E5=85=85?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 "\346\216\250\350\215\220/Recall.md" | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git "a/\346\216\250\350\215\220/Recall.md" "b/\346\216\250\350\215\220/Recall.md"
index a9b9f9d..745cdab 100644
--- "a/\346\216\250\350\215\220/Recall.md"
+++ "b/\346\216\250\350\215\220/Recall.md"
@@ -28,4 +28,4 @@
     - 第二步：用cls预测neg，得到预测prob>0.5的所有neg对应的feature和prob；
     - 第三步：把第二步得到的feature喂给cls，重新train；
     - 第四步：把第二步和第三步迭代几次；
-- 《Training Region-based Object Detectors with Online Hard Example Mining提出，用loss找出hard example，bp的时候只回传这些hard example的weight更新
\ No newline at end of file
+- 《Training Region-based Object Detectors with Online Hard Example Mining》提出，用loss找出hard example，bp的时候只回传这些hard example的weight更新
\ No newline at end of file