Sudipta Sinha’s Post

We have released code for our paper, Improved Scene Landmark Detection for Camera Localization, to be presented at 3DV 2024. This paper proposes improvements over our prior work [1]. Our new approach is competitive with the state-of-the-art on the INDOOR-6 dataset but runs significantly faster. This was a collaboration with Tien Do while he was at Microsoft. Paper: https://lnkd.in/gWsHNHHw Code: https://lnkd.in/gSW2yFez [1] Do et al. 2022, Learning to Detect Scene Landmarks for Camera Localization, in CVPR. https://lnkd.in/gYp4aGUw

Tools for smoother transitioning be your own translator.

🚀 BIG NEWS ALERT! 🚀 Ever wish you could understand all your customers without having to perfect a new language every week? Well, wish no more! 😎 Dubit is coming soon, and it's about to make “Lost in Translation” a thing of the past. 🌍💬 We’re talking real-time language conversion that’s faster than your morning coffee kicking in. ☕⚡ Ready to talk to the world in their language (without Google Translate)? Stay tuned, we’ve got some magic coming your way! 🪄✨ #DubitLaunch #NoMoreLostInTranslation #RealTimeLanguage #TechMagic #BreakingLanguageBarriers #CommunicationRevolution #BusinessUpgrade

This work presents a #wireless #localization method that operates on #self-#supervised and #unlabeled #channel #estimates. Their self-supervising method learns #general-#purpose #channel #features #robust to fading and system impairments. Learned representations are easily transferable to new environments and ready to use for other wireless downstream tasks. To the best of their knowledge, the proposed method is the first #joint-#embedding self-supervised approach to forsake the dependency on #contrastive #channel #estimates. Their approach outperforms #fully-#supervised techniques in #small #data #regimes under #fine-#tuning and, in some cases, linear evaluation. They assess the performance in centralized and distributed massive #multiple-#input #multiple-#output (#MIMO) systems for multiple datasets. ----Artan Salihu, Markus Rupp, Stefan Schwarz More details can be found at this link: https://lnkd.in/gkRvksea

Pre-translation vs. direct inference in multilingual LLM applications | Google A comprehensive evaluation comparing pre-translation with direct inference of PaLM2 on multilingual tasks, demonstrating its improved performance using direct inference in the source language, compared to pre-translation to English. PaLM2 models do not need pre-translation to excel in multilingual tasks, as demonstrated in a comprehensive evaluation comparing direct inference with pre-translation. 👉 https://lnkd.in/dvRKdAGH #machinelearning

Struggling to find native language speakers for your operation? Unable to talk to your customers in their language? Livechat+™ enables your teams to handle chat conversations in 100+ languages instantly and it’s up to 3 times more accurate than most translation tools. Quick to implement, and requiring little integration, ROI can be generated from day one of going live. Learn more: https://lnkd.in/ebZYtBkD Digital Language Partners #Digitallanguagepartners #BilingualSupport #CustomerService #MultilingualSolutions #translation

Introduction to Google #Translate Learn #GoogleTranslate tips and tricks #LanguageLearning with @Google #Google Translate Features Fun with Google Translate. Why does Google Translate is successful for #MachineLearning #ml #translation #technology #communication #multilingual Be a proud user of Google Translate.

〜 What is Accessible Code®? 〜 Accessible Code (JP Patent No. 6947960) is a QR code-based text-to-speech tool that makes printed information more accessible to visually impaired people (VIP) in their native language. In short, it can contribute to the world without any barrier of language and visually impairment. #accessibility #sustainability

Competition Launch Alert! Unlocking Global Communication with Gemma AI language translation is a strong example of where lightweight open models shine – there are over 7,000 languages, countless cultural nuances, and translations need to work quickly on smartphones. Today, Google released a new Gemma 2 2b variant fine-tuned for Japanese with training materials as a learning resource. Their goal is to empower communities worldwide to adapt Gemma to their own languages, using their deep understanding of their languages and cultures. To further support this effort, we've partnered with Google to launch the Unlocking Global Communication with Gemma competition. Participants are invited to fine-tune Gemma 2 for their languages and share their knowledge through reproducible notebooks that explore elements like language fluency, literary traditions, historical texts, and more. 💰 $150,000 Prize Pool ⏰ Entry Deadline: 1/14/25 Your work will help unlock the full potential of language AI and create a more connected and understanding world. Learn more at https://lnkd.in/eazkU7Ja 

FORE Global: building borderless predictions 🌎 FORE Predict is a predictions platform powered by people all over the globe. As we prepare to scale to a global audience, FORE Predict now features in-app translation for a number of popular languages: Spanish, Arabic, Chinese, Japanese, Portuguese, German, French, and Italian. This is just the beginning of our journey towards a borderless and inclusive prediction ecosystem: in upcoming iterations we will introduce additional languages and dynamic translation, enabling translation of user-generated content on FORE Predict. Meanwhile, additional chains, acceptance of leading assets, and new features will be deployed to drive user adoption all over the world. Let us know below where we should take FORE Predict next 🌎 Learn more: https://lnkd.in/g82F7hWf