updated readme

README.md

@@ -1,52 +1,107 @@
-# Immich Selfie Timelapse Tool
+# Immich Selfie Timelapse
 
-This tool helps create selfie timelapses from your Immich instance.  
+***Generate timelapse videos of your loved ones' portraits from your [Immich](https://immich.app/) photo library.***
-It uses the powerful machine learning features of Immich to gather all the photographs where a particular individual appears, retrieves the bounding box metadata, and automatically crops and aligns the photos.
+
+Use Immich's built-in face recognition to find all photos of a person, then runs them through a configurable processing pipeline that crops, aligns, filters, and compiles the results into a smooth timelapse video.
 
 <p align="center">
-  <img src="resources/example.jpg" alt="Example Image">
+  <img src="resources/main_view.png" alt="Main View">
 </p>
 
 <p align="center">
-  <img src="resources/example_gif.gif" alt="Example GIF">
+  <img src="resources/result.gif" alt="Timelapse example" width="40%">
 </p>
 
-
 ## Features
 
-- Automatically fetch images featuring a specified individual from your Immich instance.
+- **Face detection & cropping** - Uses Immich's face recognition metadata to locate and crop faces automatically
-- Extract bounding box metadata and crop/align photos using machine learning.
+- **Face alignment** - Aligns all photos based on eye positions for a stable, smooth timelapse
-- Discard photos with low resolution (set by threshold).
+- **Head pose filtering** - Skips non-front-facing photos using an ONNX deep learning model (DMHead)
-- Discard photos where the subject is viewed from the side.
+- **Blink detection** - Filters out photos where eyes are closed using Eye Aspect Ratio analysis
-- Adds timestamp in the filename for easy chronological ordering.
+- **Blur detection** - Discards blurry photos using gradient magnitude analysis (Sobel operator)
+- **Brightness filtering** - Removes photos that are too dark or overexposed
+- **Face resolution filtering** - Skips photos where the face is too small
+- **Photo limiting** - Caps photos per day/week/month to avoid over-representing busy periods
+- **Timestamp overlay** - Optionally overlays the date on each frame
+- **Video compilation** - Automatically compiles processed photos into an MP4 timelapse using FFmpeg
+- **Web UI** - Configure all settings and monitor progress through a built-in web interface
 
-## Docker compose setup
+<p align="center">
+  <img src="resources/processing_detail.png" alt="Processing">
+</p>
 
-```yml
+## Quick Start
+
+### Immich API key
+
+To access your immich library, this project requires an Immich API key.  
+Follow this guide to create one: https://immich.app/docs/features/command-line-interface#obtain-the-api-key
+
+Here are the features that need to be enabled:
+- server.about
+- asset.download
+- asset.read
+- person.read
+
+*Note: It is advised to store this API key in a `.env` file adjacent to your `docker-compose.yml` rather than in plain text.*
+
+### Docker Compose (recommended)
+
+```yaml
 services:
   immich-selfie-timelapse:
     image: arnaudcayrol/immich-selfie-timelapse
     container_name: immich-selfie-timelapse
     ports:
       - "5000:5000"
-    volumes:
-      - ./immich_selfie_timelapse:/app/output
     environment:
       - IMMICH_API_KEY=abcdefghijklmnopqrstuvwxyz
-      - IMMICH_BASE_URL=http://192.168.1.94:2283/api
+      - IMMICH_BASE_URL=http://your-immich-host:2283
+    volumes:
+      - ./config:/app/config
+      - ./output:/app/output
+    restart: unless-stopped
 ```
 
-Once the service is started, access the webpage to configure the tool: http://127.0.0.1:5000.  
+### Docker Run
+
+```bash
+docker run -d \
+  --name immich-selfie-timelapse \
+  -p 5000:5000 \
+  -e IMMICH_API_KEY=your-api-key-here \
+  -e IMMICH_BASE_URL=http://your_server:2283 \
+  -v ./config:/app/config \
+  -v ./output:/app/output \
+  arnaudcayrol/immich-selfie-timelapse
+```
+
+Then open **http://your_server:5000** to access the web interface.
+
+### Volumes
+
+| Path | Description |
+|---|---|
+| `/app/config` | Persisted configuration file (`config.toml`) |
+| `/app/output` | Processed images and compiled timelapse video |
+
+### Settings
+
+The default settings are quite permissive because every human being is unique.  
+Please adjust image brightness filtering, eye aspect ratio etc. for the person you are processing.
 
 <p align="center">
-  <img src="resources/webpage.jpg" alt="Example GIF">
+  <img src="resources/settings_details.png" alt="Settings">
 </p>
 
-## Additional Notes
+## Additional info
+
+- Heartfelt thanks to the Immich team and contributors for making this project possible.
+- About contribution : When I first created this project, I marked it as open to contributions. I now realize that I don't have as much time as I thought to dedicate to this project. I feel comfortable with issues being opened as it allows me to go through them at my own pace. For pull requests, I cannot guarantee a reasonable time frame for review. 
+- Thank you thomaslrg for the discussions around the project.
+- Thank you for the 200 GitHub stars !
 
-- For best looking results, I still recommand a manual pass over the final image folder to remove some outliers. The face orientation detection is not perfect at all. Hopefully someday I can get a better results.
-- Contributions and improvements are welcome.
 
 ## License
 
-This project is open source and available under the MIT License.
+This project is open source and available under the [MIT License](LICENSE).

src/config.rs

@@ -136,7 +136,7 @@ pub struct BrightnessConfig {
 impl Default for BrightnessConfig {
     fn default() -> Self {
         Self {
-            enabled: true,
+            enabled: false,
             min_brightness: 0.1,
             max_brightness: 0.9,
         }
@@ -215,7 +215,7 @@ impl Default for OutputConfig {
     fn default() -> Self {
         Self {
             size: 512,
-            keep_intermediates: true,
+            keep_intermediates: false,
         }
     }
 }
@@ -300,7 +300,7 @@ impl Default for EyeFilterConfig {
     fn default() -> Self {
         Self {
             enabled: true,
-            min_ear: 0.22,
+            min_ear: 0.12,
         }
     }
 }