The DL concepts and models that were illustrated in the previous two sections aren't just pure theory – practical applications have been implemented from them. DL excels at identifying patterns in unstructured data; most use cases are related to media such as images, sound, video, and text. Nowadays, DL is applied in a number of use case scenarios across different business domains, such as the following:
- Computer vision: A number of applications in the automotive industry, facial recognition, motion detection, and real-time threat detection
- NLP: Sentiment analysis in social media, fraud detection in finance and insurance, augmented search, and log analysis
- Medical diagnosis: Anomaly detection, pathology identification
- Search engines: Image searching
- IoT: Smart homes, predictive analysis using sensor data
- Manufacturing: Predictive maintenance
- Marketing: Recommendation engines, automated target identification
- Audio analysis: Speech recognition, voice searching, and machine translation
There are many others that are yet to come.