Stylumia Apollo Fashion Attribute Extraction

Stylumia Apollo Fashion Attribute Extraction

Stylumia's Apollo has an inbuilt prediction engine for demand forecasting of a new product in fashion by performing attribute extraction . Apollo is a customized prediction engine for the retailer, which is inspired by human visual perception through an ensemble machine learning model combining images and textual attributes powered by global fashion intelligence and reatiler's transaction data. The multi modal inputs (images, customer reviews, purchase trends, geographic info, etc) are analyzed and relevant attributes extracted for predicting the demand . This workload is used for extracting fashion attributes like sleeve length, neck type, dress length, pattern etc ., from images of E-Commerce data to better understand the trends at attribute level

*Please note that member solutions are often customizable to meet the needs of individual enterprise end users.



  • Image Segmentation and Classification
    Segmenting Retail images and classifying the Parts of the Apparel.
  • Post-season diagnosis for range correction
    Analyze past season performance at various depths of attributes, both functional and visual. This enables you to take learning from past season winners and losers to make very informed range correction/augmentation decisions


WorldwideAI Platform as a Service (AI PaaS)RetailData AnalyticsVideo Surveillance and AnalyticsCSP - Amazon Web ServicesCSP - Google CloudOn-premise (Private Cloud, Other)Intel® Xeon® Scalable ProcessorIntel® Distribution of OpenVINO™ Toolkit powered by oneAPIIntel® Optimization for PyTorch*Models can be trained - data input only requiredLinux ResNet50Deep Learning