Sai Venkatesh Ramesh


I am currently working as an Individual Contributor at Motorq Labs to build their connected vehicles analytics platform.
I previously worked as a Software Development Engineer at Samsung Digital E-Commerse in their Global Platform V2 product.
I also worked as a Research Assistant in Product Labs, IIIT-Hyderabad.
Previously, I worked as a Intern at Glosys Technology Solutions Pvt Ltd.
I graduated with a B.Eng in Computer Science and Engineering from Anna University.
I enjoy competitive programming, reading tech, political articles and cycling.

news

Sep 16, 2021 Won the Special Prize for our education tool ‘Zephyr’ at the Hack Of Pie Competition
Mar 5, 2021 Idea on “Virtual Learning” has qualified for the OpenCV AI Comp. 2021 finals.
Jan 26, 2021 Article on FairMOT : Object Detection published in Analytics Vidhya.
Jan 23, 2021 TrackJectory code available in GitHub :)
Dec 26, 2020 Won Samsung GMC Hackathon 2020.

selected publications

  1. AV
    FairMOT : Multi-Object Tracking.
    R Sai Venkatesh,

    In Analytics Vidhya 2021.

    Code Link
  2. MIKE
    GlosysIC Framework: Transformer for Image Captioning with Sequential Attention.
    Srinivasan Thanukrishnan, R Sai Venkatesh, and Vijay Vignesh Prasad Rao

    In Lecture Notes in Computer Science, Springer 2020.

    Code Link
    Over the past decade, the field of Image captioning has witnessed a lot of intensive research interests. This paper proposes “GlosysIC Framework: Transformer for Image Captioning with Sequential Attention” to build a novel framework that harnesses the combination of Convolutional Neural Network (CNN) to encode image and transformer to generate sentences. Compared to the existing image captioning approaches, GlosysIC framework serializes the Multi head attention modules with the image representations. Furthermore, we present GlosysIC architectural framework encompassing multiple CNN architectures and attention based transformer for generating effective descriptions of images. The proposed system was exhaustively trained on the benchmark MSCOCO image captioning dataset using RTX 2060 GPU and V100 GPU from Google Cloud Platform in terms of PyTorch Deep Learning library. Experimental results illustrate that GlosysIC significantly outperforms the previous state-of-the-art models.
  3. AV
    Faster R-CNN : Object Detection.
    R Sai Venkatesh,

    In Analytics Vidhya 2020.

    Code Link