diff --git a/slides/05_conv_nets_2/index.html b/slides/05_conv_nets_2/index.html
index 7fdd3d0..adfc0f2 100644
--- a/slides/05_conv_nets_2/index.html
+++ b/slides/05_conv_nets_2/index.html
@@ -326,7 +326,7 @@
 Region proposal input is a fully convolutional network: shares weights across spatial dimensions
 
 ---
-## State-of-the-art
+## Measuring performance
 
 .center[
           <img src="images/sotaresults2.png" style="width: 720px;" />
@@ -342,25 +342,6 @@
 
 - AP @0.5 for class "cat": average precision for the class, where $IoU(box^{pred}, box^{true}) > 0.5$
 
----
-## State-of-the-art - fast methods
-
-.center[
-          <img src="images/sotaresults.png" style="width: 450px;" />
-]
-
-Measures: **mAP** and Frames per Second **FPS**
-
-.footnote.small[
-Zeming Li et al. Light-Head R-CNN: In Defense of Two-Stage Object Detector 2017
-]
-
---
-
-- Mask RCNN, light-head R-CNN for best accuracy
-- Yolo, SSD, Light-Head R-CNN for fast inference
-
-
 ---
 ## State-of-the-art