Detecting Semantic Concepts from Video Using Temporal Gradients and Audio Classification