We improved ad monetization with a 20% increase in Click Through Rate (CTR) using our AI/ML-based Intelligent Ad Advisor tool for a media giant
- Extracted frame from Video contents based on similarity score indexed by time, followed by object detection to extract features, where we trained deep convolution neural net over Yolo V3 using transfer learning.
- Applied pretrained weights of Deep Speech to extract audio features from video and advertisement content.
- Combined the audio and video features and calculated the similarity score/ spot scores using BERT, which provides best spot in content to put and an advertisement.
- Developed a UI to cross promote commercials and other shows.
- We created an ad-advisor solution that enables context-based choice and placement of commercials and trailers in streaming videos
- Our application’s AI engine combines object detection models with audio-based image-captioning models to establish a syntactic and semantic context for every frame in a video
BENEFITS AND OUTCOMES
- Our solution is an AI-driven SaaS to enable contextual ad placements by adding the layer of the relevance of an Ad to the content of the shows and movies on which it is aired, this enabled our client to register a 20% increase in CTR