From 68d484d85e8b9ffe1f180fd438b5b2399a51a748 Mon Sep 17 00:00:00 2001 From: ameyawagh Date: Sun, 3 Jun 2018 10:49:39 -0400 Subject: [PATCH] loss function formatted in README --- README.md | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/README.md b/README.md index 39a2046..56f25f4 100644 --- a/README.md +++ b/README.md @@ -26,9 +26,9 @@ The network heads consists of the Mask branch which predicts the mask and a clas | Feature Pyramid network with Resnet | different head architecture with and without FPN | The loss function consists of 3 losses *L = Lclass + Lbox + Lmask* where - - `Lclass` uses log loss for true classes - - `Lbox` uses smoothL1 loss defined in [Fast RCNN] - - `Lmask` uses average binary cross entropy loss + - *Lclass* uses log loss for true classes + - *Lbox* uses smoothL1 loss defined in [Fast RCNN] + - *Lmask* uses average binary cross entropy loss The masks are predicted by a [Fully Connected Network](https://arxiv.org/pdf/1605.06211.pdf) for each RoI. This maintains the mxm dimension for each mask and thus for each instance of the object we get distinct masks.