Update README.md

README.md

@@ -7,10 +7,13 @@ It combines global attention-based reasoning with patch-level self-consistency a
 
 The model is designed to indicate **where anomalies occur** and **how strongly they deviate from expected visual consistency**, supporting explainable image analysis and integrity assessment.
 
+Paper link: *to be added upon publication*
 ---
 
 ## Architecture Overview
 
+![VAAS Methodology](./methodology.png)
+
 VAAS consists of two complementary components:
 
 - **Global Attention Module (Fx)**  
@@ -44,24 +47,40 @@ This release corresponds to:
 Future releases will scale training data size, include cross-dataset evaluation, and explore model compression.
 
 ---
-
 ## Intended Use
 
 This model can be used for:
 
-- Image anomaly detection
-- Visual integrity assessment
-- Explainable inspection of irregular regions
-- Research on attention-based anomaly scoring
-- Prototyping anomaly-aware vision systems
+* Image anomaly detection
+* Visual integrity assessment
+* Explainable inspection of irregular regions
+* Research on attention-based anomaly scoring
+* Prototyping anomaly-aware vision systems
 
-It supports both **CPU-only inference** & **GPU-only inference** , though GPU is recommended for faster processing.
+It supports **CPU-only inference** and **GPU-accelerated inference**.
+GPU usage is recommended for faster processing but is not required.
 
 ---
 
-## Usage
+## Installation
+
+VAAS is distributed as a lightweight inference library.
+Users must install **PyTorch** separately to match their system (CPU or GPU).
+
+### 1. Install PyTorch
+
+Follow the official PyTorch installation guide for your platform:
+[https://pytorch.org/get-started/locally/](https://pytorch.org/get-started/locally/)
+
+### 2. Install VAAS
+
+```bash
+pip install vaas
+```
 
-### Load the pipeline
+---
+
+## Usage
 
 ```python
 from vaas.inference.pipeline import VAASPipeline
@@ -93,33 +112,45 @@ anomaly_map = result["anomaly_map"]
 
 ---
 
+## Model Files
+
+This repository contains:
+
+* `px_model.pth` – Patch-level SegFormer model weights
+* `ref_stats.pth` – Reference statistics for anomaly normalization
+* `config.json` – Model configuration metadata
+
+The Vision Transformer backbone is loaded programmatically during inference.
+
+---
+
 ## Training Data
 
-The model was trained on a reproducible 10% subset of DF2023.
-The exact filenames used for training are released to support experiment reproducibility.
+The model was trained on a reproducible 10% subset of the DF2023 dataset.
+The exact filenames used for training are released to support experimental reproducibility.
 
 ---
 
 ## Limitations
 
-- Trained on a subset of a single dataset
-- Does not classify anomaly types
-- Performance may degrade on out-of-distribution imagery
+* Trained on a subset of a single dataset
+* Does not classify anomaly types
+* Performance may degrade on out-of-distribution imagery
 
-Users are encouraged to fine-tune or retrain for domain-specific applications.
+Users are encouraged to fine-tune or retrain the model for domain-specific applications.
 
 ---
 
 ## Ethical Considerations
 
 VAAS is intended for research and inspection purposes.
-It should not be used as a standalone decision-making system in high-stakes settings.
+It should not be used as a standalone decision-making system in high-stakes or sensitive applications without human oversight.
 
 ---
 
 ## Citation
 
-If you use this model, please cite:
+Please cite the following work (citation will be updated upon publication):
 
 ```
 Bamigbade, O., Scanlon, M., Sheppard, J.
@@ -137,5 +168,5 @@ MIT License
 
 ## Maintainers
 
-OBA-Research  
-https://huggingface.co/OBA-Research
+**OBA-Research**
+[https://huggingface.co/OBA-Research](https://huggingface.co/OBA-Research)