Icon--AD-black-48x48Icon--address-consumer-data-black-48x48Icon--appointment-black-48x48Icon--back-left-black-48x48Icon--calendar-black-48x48Icon--Checkbox-checkIcon--clock-black-48x48Icon--close-black-48x48Icon--compare-black-48x48Icon--confirmation-black-48x48Icon--dealer-details-black-48x48Icon--delete-black-48x48Icon--delivery-black-48x48Icon--down-black-48x48Icon--download-black-48x48Ic-OverlayAlertIcon--externallink-black-48x48Icon-Filledforward-right_adjustedIcon--grid-view-black-48x48Icon--info-i-black-48x48Icon--Less-minimize-black-48x48Icon-FilledIcon--List-Check-blackIcon--List-Cross-blackIcon--list-view-mobile-black-48x48Icon--list-view-black-48x48Icon--More-Maximize-black-48x48Icon--my-product-black-48x48Icon--newsletter-black-48x48Icon--payment-black-48x48Icon--print-black-48x48Icon--promotion-black-48x48Icon--registration-black-48x48Icon--Reset-black-48x48share-circle1Icon--share-black-48x48Icon--shopping-cart-black-48x48Icon--start-play-black-48x48Icon--store-locator-black-48x48Ic-OverlayAlertIcon--summary-black-48x48tumblrIcon-FilledvineIc-OverlayAlertwhishlist
Bosch Center for Artificial Intelligence
Our Research Field

Large-Scale AI and Deep Learning

What Motivates Us

Most advances in novel machine learning algorithms are developed in an academic setting and based on smaller datasets. As a result, these algorithms are not ready for production use on real-world datasets. Scaling so far has been restricted to a select few industrial settings. To efficiently deploy AI at scale, we need to develop distributed machine learning algorithms and pipelines, and work on techniques to efficiently explore multidimensional hyper-parameter spaces.

Our Approach

One way to scale training of models is through multi-task architectures: these architectures exploit the correlation between several tasks to improve performance over individual tasks. As key enabler for wider applicability, multi-task architectures need to be able to learn from noisy, unlabeled or alternatively labeled data sets. Further, we investigate combination of distributed computing paradigms with techniques for efficient hyper-parameter optimization. To scale data preprocessing, we focus on techniques combining multiple sources of data to provide ground truth labels.

Application

Techniques which are able to effectively explore multi-dimensional hyper-parameter spaces are a key enabler across domains. At Bosch, we focus on deployment of these techniques for semantic search capabilities on large-scale video databases. Our research aims to provide accurate frame-level and scene-level tags for video sequences, enabling efficient search of desired scenes.

Read more

  • Boch BCAI - ICLR

    Zhang et al.

    "Deep learning on symbolic representations for large-scale heterogeneous time-series event..."
    • Authors: Shengdong Zhang, Soheil Bahrampour, Naveen Ramakrishnan
    • Published in IEEE Xplore in 2017
  • Bosch BCAI - Publication - ESANN

    Dhar et al.

    "Universum Learning for Multiclass SVM"
    • Authors: Sauptik Dhar, Naveen Ramakrishnan, Vladimir Cherkassky, Mohak Shah
    • Published in SVM in 2016
  • Bosch BCAI - ADDM

    Dhar et al.

    "ADMM based Scalable Machine Learning on Spark"
    • Authors: Sauptik Dhar, Congrui Yi Naveen Ramakrishnan, Mohak Shah
    • Published in ADMM in 2015
  • Bosch BCAI - Publication - ICLR

    Heit et al.

    "An architecture for the deployment of statistical models for the big data era"
    • Authors: Juergen Heit, Jiayi Liu, Mohak Shah
    • Published in Big Data in 2016