Fig. 4

Diagram of fusion architecture that learns jointly from VS_IMG and DFF. (a) VS_IMG a 224X224X3 image obtained by overlaying GASF-GADF-MTF (b,c) ResNet50 pre-trained model is then used to extract features from VS_IMG (d) these flattened features are then dimensionally reduced to 3036 using PCA (e) DFF for same molecule obtained from RDKit (f) DFF are dimensionally reduced using PCA (g) PCA reduced features from (d), (f) are concatenated and later classified using CSMLP.