diff --git a/README.md b/README.md
index e554542efd11ac70e88a0487bc5696cf84ff11e0..81bd34a15f448c844cda9733e213c8506e2c7607 100644
--- a/README.md
+++ b/README.md
@@ -3,6 +3,15 @@ Class imagej_tiff to read multilayer tiff files and parse tags
 * layers are stacked along depth (think RGB)
 * parse imagej generated tags (50838 and 50839)
 
+# More info
+* Presentation for CVPR2018: [Elphel_TP-CNN_slides.pdf](https://community.elphel.com/files/presentations/Elphel_TP-CNN_slides.pdf)
+* [TIFF Image stacks for Machine Learning](https://wiki.elphel.com/wiki/Tiff_file_format_for_pre-processed_quad-stereo_sets#TIFF_image_stacks_for_ML)
+
+# Samples
+* [models/all/state_street/1527256815_150165/v01/ml/](https://community.elphel.com/3d+biquad/models/all/state_street/1527256815_150165/v01/ml/)
+or
+* go to [3d+biquad](https://community.elphel.com/3d+biquad/), open individual models and hit the light green button to ‘Download source files for ml’
+
 # Dependencies
 * Python 3.5.2 (not strict)
 * Pillow 5.1.0+ (strict)