(DOCS): add HF Space README with usage, roadmap, contributors, citation and links

README.md

@@ -2,174 +2,81 @@
 title: AI Polymer Classification
 emoji: 🔬
 colorFrom: indigo
-colorTo: green
+colorTo: teal
 sdk: streamlit
 app_file: app.py
 pinned: false
 license: apache-2.0
 ---
+## AI-Driven Polymer Aging Prediction and Classification (v0.1)
 
-# 🔬 AI-Driven Polymer Aging Prediction and Classification System
+This web application classifies the degradation state of polymers using Raman spectroscopy and deep learning.
 
-A research project developed as part of AIRE 2025. This system applies deep learning to spectral data to classify polymer aging a critical proxy for recyclability using a fully reproducible and modular ML pipeline.
+It was developed as part of the AIRE 2025 internship project at the Imageomics Institute and demonstrates a prototype pipeline for evaluating multiple convolutional neural networks (CNNs) on spectral data.
 
-The broader research vision is a multi-modal evaluation platform, benchmarking not only Raman spectra but also image-based models and FTIR spectral data, ensuring reproducibility, extensibility, and scientific rigor.
-
----
-
-## 🎯 Project Objective
-
-- Build a validated machine learning system for classifying polymer spectra (predict degradation levels as a proxy for recyclability)
-- Evaluate and compare multiple CNN architectures, beginning with Figure2CNN and ResNet variants, and expand to additional trained models.
-- Ensure scientific reproducibility through structured diaignostics and artifact control
-- Support sustainability and circular materials research through spectrum-based classification.
-
-  **Reference (for Figure2CNN baseline):**
-
-  > Neo, E.R.K., Low, J.S.C., Goodship, V., Debattista, K. (2023).
-  > Deep learning for chemometric analysis of plastic spectral data from infrared and Raman databases.
-  > Resources, Conservation & Recycling, 188, 106718.
-  > https://doi.org/10.1016/j.resconrec.2022.106718
 ---
 
-## 🧠 Model Architectures
+## 🧪 Current Scope
 
-| Model| Description |
-|------|-------------|
-| `Figure2CNN`  | Baseline model from literature |
-| `ResNet1D`    | Deeper candidate model with skip connections |
-| `ResNet18Vision` | Image-focused CNN architecture, retrained on polymer dataset (roadmap) |
-
-  Future expansions will add additional trained CNNs, supporting direct benchmarking and comparative reporting.
+- 🔬 **Modality**: Raman spectroscopy (.txt)
+- 🧠 **Model**: Figure2CNN (baseline)
+- 📊 **Task**: Binary classification — Stable vs Weathered polymers
+- 🛠️ **Architecture**: PyTorch + Streamlit
 
 ---
 
-## 📁 Project Structure (Cleaned and Current)
-
-```text
-ml-polymer-recycling/
-├── datasets/
-├── models/           # Model architectures
-├── scripts/          # Training, inference, utilities
-├── outputs/          # Artifacts: models, logs, plots
-├── docs/             # Documentation & reports
-└── environment.yml   # (local) Conda execution environment
-```
+## 🚧 Roadmap
 
-![ml-polymer-gitdiagram-0](https://github.com/user-attachments/assets/bb5d93dc-7ab9-4259-8513-fb680ae59d64)
+- [x] Inference from Raman `.txt` files
+- [x] Model selection (Figure2CNN, ResNet1D)
+- [ ] Add more trained CNNs for comparison
+- [ ] FTIR support (modular integration planned)
+- [ ] Image-based inference (future modality)
 
 ---
 
-## ✅ Current Status
-
-| Track     | Status               | Test Accuracy |
-|-----------|----------------------|----------------|
-| **Raman** | ✅ Active & validated  | **87.81% ± 7.59%** |
-| **Image**  | 🚧 Planned Expansion | N/A |
-| **FTIR**  | ⏸️ Deferred/Modularized | N/A |
-
-## 🔬 Key Features
-
-- ✅ 10-Fold Stratified Cross-Validation
-- ✅ CLI Training: `train_model.py`
-- ✅ CLI Inference `run_inference.py`
-- ✅ Output artifact naming per model
-- ✅ Raman-only preprocessing with baseline correction, smoothing, normalization
-- ✅ Structured diagnostics JSON (accuracies, confusion matrices)
-- ✅ Canonical validation script (`validate_pipeline.sh`) confirms reproducibility of all core components
-
----
-
-**Environments:**
-
-```bash
-# Local
-git checkout main
-conda env create -f environment.yml
-conda activate polymer_env
+## 🧭 How to Use
 
-# HPC
-git checkout hpc-main
-conda env create -f environment_hpc.yml
-conda activate polymer_env
-```
+1. Upload a Raman spectrum `.txt` file (or select a sample)
+2. Choose a model from the sidebar
+3. Run analysis
+4. View prediction, logits, and technical information
 
-## 📊 Sample Training & Inference
+Supported input:
 
-### Training (10-Fold CV)
-
-```bash
-python scripts/train_model.py --model resnet --target-len 4000 --baseline --smooth --normalize
-```
-
-### Inference (Raman)
-
-```bash
-python scripts/run_inference.py --target-len 4000 
---input datasets/rdwp/sample123.txt --model outputs/resnet_model.pth 
---output outputs/inference/prediction.txt
-```
-
-### Inference Output Example:
-
-```bash
-Predicted Label: 1 True Label: 1
-Raw Logits: [[-569.544, 427.996]]
-```
-
-### Validation Script (Raman Pipeline)
-
-```bash
-./validate_pipeline.sh
-# Runs preprocessing, training, inference, and plotting checks
-# Confirms artifact integrity and logs test results
-```
+- Plaintext `.txt` files with 1–2 columns
+- Space- or comma-separated
+- Comment lines (#) are ignored
+- Automatically resampled to 500 points
 
 ---
 
-## 📚 Dataset Resources
+## Contributors
 
-| Type  | Dataset | Source |
-|-------|---------|--------|
-| Raman | RDWP    | [A Raman database of microplastics weathered under natural environments](https://data.mendeley.com/datasets/kpygrf9fg6/1) |
+  👨‍🏫 Dr. Sanmukh Kuppannagari (Mentor)  
+  👨‍🏫 Dr. Metin Karailyan (Mentor)  
+  👨‍💻 Jaser Hasan (Author/Developer)
 
-| Datasets should be downloaded separately and placed here:
+## 🧠 Model Credit
 
-```bash
-datasets/
-└── rdwp/
-  ├── sample1.txt
-  ├── sample2.txt
-  └── ...
-```
+Baseline model inspired by:
 
-These files are intentionally excluded from version control via `.gitignore`
+Neo, E.R.K., Low, J.S.C., Goodship, V., Debattista, K. (2023).  
+*Deep learning for chemometric analysis of plastic spectral data from infrared and Raman databases.*  
+_Resources, Conservation & Recycling_, **188**, 106718.  
+[https://doi.org/10.1016/j.resconrec.2022.106718](https://doi.org/10.1016/j.resconrec.2022.106718)
 
 ---
 
-## 🛠 Dependencies
+## 🔗 Links
 
-- `Python 3.10+`
-- `Conda, Git`
-- `PyTorch (CPU & CUDA)`
-- `Numpy, SciPy, Pandas`
-- `Scikit-learn`
-- `Matplotlib, Seaborn`
-- `ArgParse, JSON`
+- 💻 **Live App**: [Hugging Face Space](https://huggingface.co/spaces/dev-jas/polymer-aging-ml)
+- 📂 **GitHub Repo**: [ml-polymer-recycling](https://github.com/KLab-AI3/ml-polymer-recycling)
 
----
-
-## 🧑‍🤝‍🧑 Contributors
-
-- **Dr. Sanmukh Kuppannagari** — Research Mentor
-- **Dr. Metin Karailyan** — Research Mentor
-- **Jaser H.** — AIRE 2025 Intern, Developer  
-
----
 
 ## 🎯 Strategic Expansion Objectives (Roadmap)
 
-> The roadmap defines three major expansion paths designed to broaden the system’s capabilities and impact:
+**The roadmap defines three major expansion paths designed to broaden the system’s capabilities and impact:**
 
 1. **Model Expansion: Multi-Model Dashboard**
 
@@ -203,11 +110,3 @@ These files are intentionally excluded from version control via `.gitignore`
     - **Phased Development**: Implementation details to be refined during meetings to ensure scientific rigor.
 
     This guarantees FTIR becomes a supported modality without undermining the validated Raman foundation.
-
-## 🔑 Guiding Principles
-
-- **Preserve the Raman baseline** as the reproducible ground truth
-- **Additive modularity**: Models, images, and FTIR added as clean, distinct layers rather than overwriting core functionality
-- **Transparency & reproducibility**: All expansions documented, tested, and logged with clear outputs.
-- **Future-oriented design**: Workflows structured to support ongoing collaboration and successor-safe research.
-