Multi-Scale Multi-Feature Context Modeling for Scene Recognition in the Semantic Manifold | Publicación