Ragav Sachdeva

Founder @takecareos (YC P26) | PhD @Oxford_VGG | Ex Nvidia, Google & Microsoft

[Ragav Sachdeva @RagavSachdeva]

Super excited to finally launch something I've been working on quietly for the last couple of months.

After launching our first pilots last month, we’ve signed 6 agencies (representing 200+ employees) to run their entire operations on @takecareos.

[Y Combinator @ycombinator]

.@TakeCareOS is an AI-first operating system for care agencies. It consolidates every function an agency runs on — rostering, shifts, CRM, care notes, messaging, invoicing, timesheets, and compliance — into one place, with AI agents that automate the admin work coordinators currently do by hand.

Congrats on the launch, @ragavsachdeva!

https://t.co/6EZeqcSMic

[Lele @EmanueleVivoli]

⏰ Last call!
https://t.co/Pa8NjIChXd

3 days left to submit a 2–4 page extended abstract to COMIQ @ ICCV 2025 — the workshop on AI for comics/manga.

Published work & WIP welcome; posters + select orals. Deadline: Aug 29.

Submit: https://t.co/FQSgbTwdgC 🎨📚🤖

#ICCV2025

[Ragav Sachdeva @RagavSachdeva]

🚨🚨🚨Deadline <10 days 🚨🚨🚨

Submit your extended abstracts on AI driven comic analysis to the COMIQ workshop #ICCV2025

https://t.co/zrvIB1qr4D

[Ragav Sachdeva @RagavSachdeva]

🚨 #ICCV2025 Workshop Alert! 🚨

🔥[COMIQ] Comic Intelligence Quotient: Advances and Challenges in AI-driven Comic Analysis

We’re exploring how machines interpret abstract visual storytelling media; dare I say, a true test for AGI 🫣

🔗 Pls consider submitting abstracts. Link👇

[Ragav Sachdeva @RagavSachdeva]

https://t.co/zrvIB1qr4D

[Ragav Sachdeva @RagavSachdeva]

🚨 #ICCV2025 Workshop Alert! 🚨

🔥[COMIQ] Comic Intelligence Quotient: Advances and Challenges in AI-driven Comic Analysis

We’re exploring how machines interpret abstract visual storytelling media; dare I say, a true test for AGI 🫣

🔗 Pls consider submitting abstracts. Link👇

[Lele @EmanueleVivoli]

💥 But that’s not all! We'll be in Milan for #ECCV2024 with our work, "ComiCap: A VLMs Pipeline for Dense Captioning of Comic Panels" 🚀

This paper presents a VLM-based approach to generate dense captions for comic panels (no 🚂), and a new metric !
https://t.co/XmYYyRFxBt

[Ragav Sachdeva @RagavSachdeva]

@EmanueleVivoli Super helpful thanks for compiling this!

[Ragav Sachdeva @RagavSachdeva]

@EmanueleVivoli @gyunginshin @Oxford_VGG Hi, thanks. No, this was not submitted to ECCV. I'm afraid I won't be in Milan 🥲

[Ragav Sachdeva @RagavSachdeva]

Presenting 🔮Tails Tell Tails, a follow-up to our recent CVPR paper, 🧙‍♂️The Manga Whisperer.

✍️ Joint work with @gyunginshin and Andrew Zisserman @Oxford_VGG.
📢 For model and datasets visit: https://t.co/Lfn9VG7a52

[AK @_akhaliq]

Tails Tell Tales

Chapter-Wide Manga Transcriptions with Character Names

https://t.co/Qmnoc1BoGC

Enabling engagement of manga by visually impaired individuals presents a significant challenge due to its inherently visual nature. With the goal of fostering accessibility, this paper aims to generate a dialogue transcript of a complete manga chapter, entirely automatically, with a particular emphasis on ensuring narrative consistency. This entails identifying (i) what is being said, i.e., detecting the texts on each page and classifying them into essential vs non-essential, and (ii) who is saying it, i.e., attributing each dialogue to its speaker, while ensuring the same characters are named consistently throughout the chapter. To this end, we introduce: (i) Magiv2, a model that is capable of generating high-quality chapter-wide manga transcripts with named characters and significantly higher precision in speaker diarisation over prior works; (ii) an extension of the PopManga evaluation dataset, which now includes annotations for speech-bubble tail boxes, associations of text to corresponding tails, classifications of text as essential or non-essential, and the identity for each character box; and (iii) a new character bank dataset, which comprises over 11K characters from 76 manga series, featuring 11.5K exemplar character images in total, as well as a list of chapters in which they appear.

[Ragav Sachdeva @RagavSachdeva]

@gyunginshin @Oxford_VGG Demo: https://t.co/f5MqCo7HNl

[Ragav Sachdeva @RagavSachdeva]

@tab_sebas Thanks for your interest! The v2 model is up now, with demo and dataset to follow soon.

[Ragav Sachdeva @RagavSachdeva]

@anshulnasery @CVPR 🙏🙏

[Ragav Sachdeva @RagavSachdeva]

Thank you everyone who stopped by my poster! I took some liberties with its layout and I'm glad it was kindly received. I hope to see more manga related projects @CVPR 2025 ❤️

[Anshul Nasery @anshulnasery]

@CVPR Manga Whisperer had a beautiful poster themed like a manga!

[Ragav Sachdeva @RagavSachdeva]

🚀🚀 Magi has hit 10,000+ downloads on🤗Hugging Face. Truly unexpected!

[Ragav Sachdeva @RagavSachdeva]

📢 The Manga Whisperer will appear at #CVPR2024 @CVPR
🚀 I really didn't expect 1600+ downloads on 🤗Hugging Face in the last two months. See you all in Seattle! 🫰

[Ragav Sachdeva @RagavSachdeva]

@9chnee @CVPR Whoops, didn't realise they weren't. Thanks for flagging and fixed!

[Ragav Sachdeva @RagavSachdeva]

📢 The Manga Whisperer will appear at #CVPR2024 @CVPR
🚀 I really didn't expect 1600+ downloads on 🤗Hugging Face in the last two months. See you all in Seattle! 🫰

[Ragav Sachdeva @RagavSachdeva]

📃 The Manga Whisperer: Automatically Generating Transcriptions for Comics
✍️ Ragav Sachdeva, Andrew Zisserman @Oxford_VGG

📕 arXiv: https://t.co/btfjex9EX9
💻 github: https://t.co/Lfn9VG6Cfu
🤗try it yourself: https://t.co/Teg8FKgoI4

[Ragav Sachdeva @RagavSachdeva]

@_PeterTakahashi @Oxford_VGG Hi. Thanks. I made it to make manga accessible to blind/visually-impaired people:)

[Ragav Sachdeva @RagavSachdeva]

📃 The Manga Whisperer: Automatically Generating Transcriptions for Comics
✍️ Ragav Sachdeva, Andrew Zisserman @Oxford_VGG

📕 arXiv: https://t.co/btfjex9EX9
💻 github: https://t.co/Lfn9VG6Cfu
🤗try it yourself: https://t.co/Teg8FKgoI4

[Ragav Sachdeva @RagavSachdeva]

Lots of room for improvement but looking forward to present our recent work at #OpenSUN3D #ICCV23

Input: 2 RGB images of an "in the wild" scene
Output: Class-agnostic changed region predictions in **both** the images

[Dmytro Mishkin 🇺🇦 @ducha_aiki]

The Change You Want to See (Now in 3D)

@RagavSachdeva, Andrew Zisserman

tl;dr: wide baseline change detection via image matching and monodepth.

https://t.co/GRu9aWhOhJ

[Ragav Sachdeva @RagavSachdeva]

@shariq_farooq This is really a shame. I used ZoeDepth for a project I was working on and I was pleasantly surprised by how good the model is. Thanks for the amazing work and making it easily accessible!