Update README.md

README.md

@@ -22,11 +22,14 @@ Each sample contains **aligned vision and time-series modalities**, enabling joi
 
 - **Modality 1 (Vision)**: Spectrogram images  
   - Format: PNG  
-  - Resolution: **224 × 224 × 3**
+  - Resolution: **224 × 224 × 3**  
+  - One spectrogram image per sample
 
 - **Modality 2 (Time Series)**: Network KPIs  
   - Format: CSV / structured sequences  
-  - Multivariate time series with features:
+  - **Fixed-length multivariate time series**
+  - **256 measurements per KPI feature, per sample**
+  - KPI features:
     - `Time`
     - `Latency`
     - `Jitter`
@@ -36,18 +39,29 @@ Each sample contains **aligned vision and time-series modalities**, enabling joi
 
 - **Labels**: Benign traffic and multiple jamming attack types
 
-The dataset is designed to be **federated-learning friendly**, with samples optionally grouped by **client/device identifiers** to emulate decentralized data ownership.
+Each spectrogram image and KPI time series correspond to the **same time window**.
 
 ---
 
-## Dataset Statistics
+## Dataset Schema
 
-*(Replace X, Y, Z with the final numbers.)*
+Each dataset sample contains:
 
-| Subset | Train | Test | Total |
-|------|------:|-----:|------:|
-| FedJam-Sample | X | Y | Z |
-| **Total** | **X** | **Y** | **Z** |
+- **`spectrogram`** *(Image)*  
+  Spectrogram image of shape **224 × 224 × 3**.
+
+- **`kpis`** *(sequence / array)*  
+  Multivariate time series of shape **[256, F]**, where:
+  - `256` is the number of temporal measurements
+  - `F` is the number of KPI features  
+    (`Time`, `Latency`, `Jitter`, `Packet Loss Count`, `Noise`, `SNR`)
+
+- **`label`** *(integer)*  
+  Encoded class label:
+  - `0`: Benign  
+  - `1`: Jamming type A  
+  - `2`: Jamming type B  
+  - `3`: Jamming type C  
 
 ---
 
@@ -57,7 +71,7 @@ The dataset is designed to be **federated-learning friendly**, with samples opti
 from datasets import load_dataset
 
 # Load the FedJam dataset
-dataset = load_dataset("YOUR_ORG_OR_USERNAME/fedjam")
+dataset = load_dataset("panitsasi/FedJam")
 
 # Access splits
 train_data = dataset["train"]
@@ -65,4 +79,21 @@ test_data  = dataset["test"]
 
 # Access a sample
 sample = train_data[0]
+
+# Modalities
+spectrogram = sample["spectrogram"]
+kpis = sample["kpis"]
+label = sample["label"]
+
 print(sample.keys())
+
+
+## Citation
+
+If you use this dataset, please cite:
+
+I. Panitsas, I. Ofeidis, and L. Tassiulas,
+“FedJam: Multimodal Federated Learning Framework for Jamming Detection,”
+arXiv:2508.09369 [cs.NI], 2025. doi:10.48550/arXiv.2508.09369.
+
+---