Update README.md

Author: danielegrattarola

README.md

@@ -1,53 +1,67 @@
 # Spectral Clustering with Graph Neural Networks for Graph Pooling
-Experimental results obtained with the MinCutPool layer as presented in the 2020 ICML paper [Spectral Clustering with Graph Neural Networks for Graph Pooling](https://arxiv.org/abs/1907.00481)
 
 <img src="./figs/mincutpool.png" width="400" height="200">
 
-This repository is based on [MinCutPool implementation](https://graphneural.network/layers/pooling/#mincutpool) provided by [Spektral](https://graphneural.network/), the Keras/TensorFlow library for Graph Neural Networks. A Pytorch implementation of MinCutPool is also available in the [Pytorch Geometric](https://pytorch-geometric.readthedocs.io/en/latest/modules/nn.html#torch_geometric.nn.dense.mincut_pool.dense_mincut_pool) library.
+This code reproduces the experimental results obtained with the MinCutPool layer 
+as presented in the ICML 2020 paper 
 
-**Required libraries**
+[Spectral Clustering with Graph Neural Networks for Graph Pooling](https://arxiv.org/abs/1907.00481)
+F. M. Bianchi, D. Grattarola, C. Alippi
 
-- spektral (tested on v.1.2)
-- tensorflow (tested on v.1.14)
-- keras (tested on v.2.2.4)
-- scikit-learn (tested on v.0.22.1)
-- scikit-image (tested on v.0.16.2)
-- networkx (tested on v.2.4)
-- pygsp (tested on v.0.5.1 )
-- pandas 
-- scipy 
-- tqdm
-- numpy
-- matplotlib 
+The official implementation of the MinCutPool layer can be found in 
+[Spektral](https://graphneural.network/layers/pooling/#mincutpool). 
+
+An implementation of MinCutPool for PyTorch is also available in 
+[Pytorch Geometric](https://pytorch-geometric.readthedocs.io/en/latest/modules/nn.html#torch_geometric.nn.dense.mincut_pool.dense_mincut_pool).
+
+## Setup
+
+The code is based on Python 3.5, TensorFlow 1.15, and Spektral 0.1.2. 
+All required libraries are listed in `requirements.txt` and can be installed with
+
+```bash
+pip install -r requirements.txt
+``` 
 
 ## Image segmentation
 
 <img src="./figs/overseg_and_rag.png" width="700" height="150">
 
-Run [Segmentation.py](https://github.com/FilippoMB/Spectral-Clustering-with-Graph-Neural-Networks-for-Graph-Pooling/blob/master/Segmentation.py) to perform hyper-segmentation, generate a Region Adjacency Graph from the resulting segments, and then cluster the nodes of the RAG graph by means of a GNN equipped with the MinCutPool layer.
+Run [Segmentation.py](https://github.com/FilippoMB/Spectral-Clustering-with-Graph-Neural-Networks-for-Graph-Pooling/blob/master/Segmentation.py) 
+to perform hyper-segmentation, generate a Region Adjacency Graph from the 
+resulting segments, and then cluster the nodes of the RAG graph with the 
+MinCutPool layer.
 
 ## Clustering
 
 <img src="./figs/clustering_stats.png" width="600" height="250">
 
-Run [Clustering.py](https://github.com/FilippoMB/Spectral-Clustering-with-Graph-Neural-Networks-for-Graph-Pooling/blob/master/Clustering.py) to cluster the nodes of a citation network. The datasets ````cora````, ````citeseer````, and ````pubmed```` can be selected.
-Resutls are provided in terms of homogeneity score, completeness score, and normalized mutual information (v-score).
+Run [Clustering.py](https://github.com/FilippoMB/Spectral-Clustering-with-Graph-Neural-Networks-for-Graph-Pooling/blob/master/Clustering.py) 
+to cluster the nodes of a citation network. The datasets `cora`, `citeseer`, and 
+`pubmed` can be selected.
+Results are provided in terms of homogeneity score, completeness score, and 
+normalized mutual information (v-score).
 
 ## Autoencoder
 
 <img src="./figs/ae_ring.png" width="400" height="200">
 <img src="./figs/ae_grid.png" width="400" height="200">
 
-Run [Autoencoder.py](https://github.com/FilippoMB/Spectral-Clustering-with-Graph-Neural-Networks-for-Graph-Pooling/blob/master/Autoencoder.py) to compute the reconstruction from an Autoencoder with bottleneck. It is possible to switch between the ````ring```` and ````grid````, but also the other [point-clouds datasets](https://pygsp.readthedocs.io/en/stable/reference/graphs.html?highlight=bunny#graph-models) from the [PyGSP](https://pygsp.readthedocs.io/en/stable/index.html) library are supported. Results are provided in terms of Mean Squared Error.
+Run [Autoencoder.py](https://github.com/FilippoMB/Spectral-Clustering-with-Graph-Neural-Networks-for-Graph-Pooling/blob/master/Autoencoder.py) 
+to train an autoencoder with bottleneck and compute the reconstructed graph. It 
+is possible to switch between the `ring` and `grid` graphs, but also any other 
+[point clouds](https://pygsp.readthedocs.io/en/stable/reference/graphs.html?highlight=bunny#graph-models) 
+from the [PyGSP](https://pygsp.readthedocs.io/en/stable/index.html) library 
+are supported. Results are provided in terms of the Mean Squared Error.
 
 ## Graph Classification
 
 Run [Classification.py](https://github.com/FilippoMB/Spectral-Clustering-with-Graph-Neural-Networks-for-Graph-Pooling/blob/master/Classification.py) to train a graph classifier. Additional classification datasets are available [here](https://chrsmrrs.github.io/datasets/) (drop them in ````data/classification/````) and [here](https://github.com/FilippoMB/Benchmark_dataset_for_graph_classification) (drop them in ````data/````).
-Results are provided in terms of classification accuray averaged over 10 folds.
+Results are provided in terms of classification accuracy averaged over 10 runs.
 
 ## Citation
 
-Please, cite the original paper if you are using MinCutPool in your reasearch
+Please, cite the original paper if you are using MinCutPool in your research
 
 	@inproceedings{bianchi2020mincutpool,
         title={Spectral Clustering with Graph Neural Networks for Graph Pooling},
@@ -58,6 +72,6 @@ Please, cite the original paper if you are using MinCutPool in your reasearch
         organization={ACM}
     }
 
-
 ## License
+
 The code is released under the MIT License. See the attached LICENSE file.