clean up docs

doc/doc_build_and_repository_info.md

@@ -1,15 +1,16 @@
-### **CPU only mode**
+## Build and Repository Info
 
-CPU mode enabled by arg --cpu-only for all stages. Follow requirements-cpu.txt to install req packages.
-Do not use DLIB extractor in CPU mode, it's too slow.
-Only H64 or SAE (with low settings) models reasonable to train on home CPU.
+DeepFaceLab officially supports Windows-only. Linux and OS X are untested.
 
-### **Build info**
 
-dlib==19.10.0 from pip compiled without CUDA. Therefore you have to compile DLIB manually, orelse use MT extractor only.
+#### **Installing dlib on Windows**
 
-Command line example for windows: `python setup.py install -G "Visual Studio 14 2015" --yes DLIB_USE_CUDA`
+The version of `dlib` in pip is compiled without CUDA support. Therefore you have to compile it manually in order to use the `dlib` face extractor.
 
-### Mac/linux/docker script support.
+To build this on Windows run following command:: `python setup.py install -G "Visual Studio 14 2015" --yes DLIB_USE_CUDA`
 
-If you want to support mac/linux/docker - create fork, it will be referenced here.
+#### **CPU mode**
+
+It is possible to run from script for all stages using the `--cpu-only` flag. To run from script, install the separate dependencies for CPU mode using `pip -r requirements-cpu.txt`.
+
+Please note that extraction and training will take much long without a GPU and performance will greatly suffer without one. In particular, do not use DLIB extractor in CPU mode, it's too slow to run without a GPU. Train only on 64px resolution models like H64 or SAE (with low settings) and the lightweight encoder.

doc/doc_features.md

@@ -1,27 +1,39 @@
 ### **Features**:
 
-- standalone zero dependencies ready to work prebuilt binary for all windows versions, see below
+- Regularly updated Windows binary containing pre-compiled dependencies, including CUDA libraries.
 
-- new models
+- New models expanding upon the original df model.
 
-- new architecture, easy to experiment with models
+- Model architecture designed with experimentation in mind.
 
-- face data embedded to JPG files
+- Face metadata embedded into extracted JPG files.
 
-- cpu mode. 8th gen Intel core CPU able to train H64 model in 2 days.
+- CPU-only mode [`--cpu-mode`]
 
-- new preview window
+- Preview window
 
-- extractor in parallel
+- Extractor and Converter can be run in parallel.
 
-- converter in parallel
+- Debug mode option for all stages: [`--debug`]
 
-- **--debug** option for all stages
+- Multiple extraction modes: MTCNN, dlib, or manual.
 
-- **MTCNN extractor** which produce less jittered aligned face than DLIBCNN, but can produce more false faces. Comparison dlib (at left) vs mtcnn on hard case:
+#### Extractor Examples
+##### MTCNN
+
+Predicts faces more uniformly than dlib, resulting in a less jittered aligned output. However, MTCNN extraction will produce more false positives.
+
+
+Comparison dlib (at left) vs mtcnn on hard case:
 ![](https://i.imgur.com/5qLiiOV.gif)
-MTCNN produces less jitter.
 
-- **Manual extractor**. You can fix missed faces manually or do full manual extract:
+- **Manual Extractor**
+
+A manual extractor is available. This extractor uses the preview GUI to allow the user to properly align detected faces. 
+
 ![](manual_extractor_0.jpg)
+
+This mode can also be used to fix incorrectly extracted faces. Manual extraction can be used to greatly improve training on face sets that are heavily obstructed. 
+
 ![Result](https://user-images.githubusercontent.com/8076202/38454756-0fa7a86c-3a7e-11e8-9065-182b4a8a7a43.gif)
+

doc/doc_prebuilt_windows_app.md

@@ -1,15 +1,16 @@
-### Prebuilt windows app:
+### **Prebuilt Windows Releases**
 
-Windows 7,8,8.1,10 zero dependency (just install/update your GeForce Drivers) prebuilt DeepFaceLab (include GPU and CPU versions) can be downloaded from 
-1) torrent https://rutracker.org/forum/viewtopic.php?p=75318742 (magnet link inside).
-2) https://mega.nz/#F!b9MzCK4B!zEAG9txu7uaRUjXz9PtBqg
+Windows builds with all dependencies included are released regularly. Only the NVIDIA GeForce display driver needs to be installed. Prebuilt DeepFaceLab, including GPU and CPU versions, can be downloaded from [Mega](https://mega.nz/#F!b9MzCK4B!zEAG9txu7uaRUjXz9PtBqg) or [BitTorrent](https://rutracker.org/forum/viewtopic.php?p=75318742).
 
-### Video tutorials for prebuilt windows app:
 
-Basic workflow: https://www.youtube.com/watch?v=K98nTNjXkq8
 
-Basic workflow (derpfakes): https://www.youtube.com/watch?v=cVcyghhmQSA
 
-How To Make DeepFakes With DeepFaceLab - An Amatuer's Guide: https://www.youtube.com/watch?v=wBax7_UWXvc
+#### Video tutorials using prebuilt windows app
 
-Manual re-extract bad dst aligned frames: https://www.youtube.com/watch?v=7z1ykVVCHhM
+* [Basic workflow](https://www.youtube.com/watch?v=K98nTNjXkq8)
+
+* [Basic workflow (thanks @derpfakes)](https://www.youtube.com/watch?v=cVcyghhmQSA)
+
+* [How To Make DeepFakes With DeepFaceLab - An Amatuer's Guide](https://www.youtube.com/watch?v=wBax7_UWXvc)
+
+* [Manual re-extract poorly aligned frames](https://www.youtube.com/watch?v=7z1ykVVCHhM)

doc/doc_ready_to_work_facesets.md

@@ -1,4 +1,6 @@
-### **Ready to work facesets**:
+### **Example Face Sets**:
+
+Faces sets for the following have been pre-extracted for experimentation,
 
 - Nicolas Cage
 - Steve Jobs
@@ -6,4 +8,4 @@
 - Elon Musk
 - Harrison Ford
 
-download from https://mega.nz/#F!y1ERHDaL!PPwg01PQZk0FhWLVo5_MaQ
+[Download from Mega](https://mega.nz/#F!y1ERHDaL!PPwg01PQZk0FhWLVo5_MaQ)

doc/doc_sort_tool.md

@@ -4,9 +4,9 @@
 
 `hist` groups images by similar content
 
-`hist-dissim` places most similar to each other images to end.
+`hist-dissim` group images by dissimilarity, placing the most similar to each other images to end.
 
-`hist-blur` sort by blur in groups of similar content
+`hist-blur` sort by a function of similar content and blur 
 
 `face-pitch` sort by face pitch direction
 
@@ -16,18 +16,22 @@
 
 `hue`
 
-`black` Places images which contains black area at end of folder. Useful to get rid of src faces which cutted by screen.
+`black` places images which contains black area at end of folder. Useful to get rid of src faces which cutted by screen.
 
 `final` sorts by yaw, blur, and hist, and leaves best 1500-1700 images.
 
-Best practice for gather src faceset from tens of thousands images:
+### **Extraction Workflow**
 
-1) `black` -> then delete faces cutted by black area at end of folder
-2) `blur` -> then delete blurred faces at end of folder
-3) `hist` -> then delete groups of similar unwanted faces and leave only target face
-4) `final` -> then delete faces occluded by obstructions
+Extraction is rarely perfect and a final pass by human eyes is typically required. This is the best way to remove unmatched or misaligned faces quickly. The sort tool included in the project greatly reduces the time and effort required to clean large sets. Like pictures will be grouped together and false positives can be quickly be identified.
 
-Best practice for dst faces:
+Suggested sort workflow for gathering cleaning face sets from very large image pools:
 
-1) delete first unsorted aligned groups of images what you can to delete. Dont touch target face mixed with others.
-2) `hist` -> then delete groups of similar and leave only target face
+1) `black` -> then delete faces with black edges at end of folder
+2) `blur` -> then delete blurred faces at end of folder, use your judgement
+3) `hist` -> then delete groups of mismatched faces, leaving only target face
+4) `final` -> then delete faces blocked by obstructions (hands, hair, etc)
+
+Suggested sort workflow for preparing and cleaning face sets from very large image pools:
+
+1) Manually delete unsorted aligned groups of images what you can to delete, ignore instances of your target face for now.
+2) `hist` -> then delete groups of mismatched faces, leaving only target face.