Skip to Main Content
Workshop Intermediate

How to build an image search engine like Google Lens using Qdrant.

Proposal status is Rejected
Session Description
  1. In this session, we shall explore how to build a system similar to Visual Similarity Search, such as Google Lens.
  2. To achieve this, we would need models trained on large datasets of images.
  3. Two prominent open source models suited for this task are google/vit-base-patch16-224-in21k and facebook/dino-vits16.
  4. These models belong to a category known as Vision Transformers (ViT).
  5. Vision Transformers (ViT) are designed to process and understand visual data.
  6. We shall also use Qdrant, An open-source vector database to store our vector embeddings of the images and also to perform the search.
  7. Please read the complete blog from the session reference to understand in depth

References

https://medium.com/gitconnected/the-big-bang-theory-of-ai-exploding-with-vectors-part-1-dc47f40e5bd8

Session Categories

FOSS

Speakers

Prathap Chandra
Prathap Chandra Lead Software Development Engineer | Freecharge DBAT - Axis Bank

I'm a 2019 CS Graduate based out of Hyderabad.

Currently working on quick overdraft product for enterprises at Freecharge DBAT - Axis Bank as a Lead SDE.

Apart from work I explore other technologies whenever I find time.

I'm an active member of Hyderabad FOSS United community. Have given two talks so far

In my leisure time I watch movies and hangout with friends