🎙️ Turn Your Voice into an Opportunity! 📲 Scan the QR code on the poster and apply today! ✨ Your Voice Can Create Real Change! #WeAreHiring #SPIRELab #IISc #PartTimeJobs #SpeechRecording #ArtificialIntelligence #AIResearch #WorkFromAnywhere

Every celebration has a voice. ❤️ From Bengali and Assamese to Odia, Kokborok and Chakma, every language carries a unique story, culture and identity. Wishing everyone a joyful and blessed Durga Puja! ✨ #ProjectVaani #LinguisticDiversity #InclusiveAI #SpireLab #ARTPARK

What happens to a voice assistant when a pressure cooker whistles mid-sentence? For most speech models, it falls apart. That’s the problem the IndoML team set for 250+ teams at the IndoML 2026 Datathon. #IndoML2026 #SpeechAI #IndicLanguages #OpenData #ResponsibleAI

Interspeech 2026 is here, and SPIRE Lab will be at ICC Sydney from 27 Sep to 1 Oct. 𝐖𝐡𝐚𝐭 𝐰𝐞'𝐥𝐥 𝐛𝐞 𝐩𝐫𝐞𝐬𝐞𝐧𝐭𝐢𝐧𝐠:

𝐖𝐞 𝐚𝐫𝐞 𝐡𝐢𝐫𝐢𝐧𝐠! | 𝐅𝐫𝐞𝐞𝐥𝐚𝐧𝐜𝐞 𝐎𝐩𝐩𝐨𝐫𝐭𝐮𝐧𝐢𝐭𝐲 📢 𝐕𝐨𝐢𝐜𝐞 𝐃𝐚𝐭𝐚 𝐐𝐮𝐚𝐥𝐢𝐭𝐲 𝐂𝐡𝐞𝐜𝐤 (𝐐𝐂) 𝐅𝐫𝐞𝐞𝐥𝐚𝐧𝐜𝐞𝐫𝐬 𝐟𝐨𝐫 𝐍𝐨𝐫𝐭𝐡 𝐄𝐚𝐬𝐭𝐞𝐫𝐧 𝐋𝐚𝐧𝐠𝐮𝐚𝐠𝐞𝐬

𝐖𝐞'𝐫𝐞 𝐡𝐢𝐫𝐢𝐧𝐠! 🚀 𝐃𝐚𝐭𝐚 𝐂𝐮𝐫𝐚𝐭𝐢𝐨𝐧 𝐀𝐬𝐬𝐨𝐜𝐢𝐚𝐭𝐞 📍 On-site @ ARTPARK, IISc Campus, Bengaluru 💼 Experience: 0–2 Years (Freshers welcome!) 📄 Engagement: Contractual

We are hiring! | Freelance Opportunity 📢 We are looking for 𝐕𝐨𝐢𝐜𝐞 𝐃𝐚𝐭𝐚 𝐐𝐮𝐚𝐥𝐢𝐭𝐲 𝐂𝐡𝐞𝐜𝐤 (𝐐𝐂) 𝐅𝐫𝐞𝐞𝐥𝐚𝐧𝐜𝐞𝐫𝐬 to support our audio data annotation and linguistic representation efforts. This is a 100% Remote (WFH) role!

𝗠𝗼𝗿𝗲 𝗹𝗮𝗻𝗴𝘂𝗮𝗴𝗲𝘀. 𝗠𝗼𝗿𝗲 𝘃𝗼𝗶𝗰𝗲𝘀. 𝗠𝗼𝗿𝗲 𝗽𝗼𝘀𝘀𝗶𝗯𝗶𝗹𝗶𝘁𝗶𝗲𝘀. What does it take for speech AI to actually hear India? It takes going past the handful of languages that most voice models have quietly settled for.

Open under an MIT licence Try it in the language you speak. Build on it. Tell us how it does, so the next version hears you better. Full conversation: youtube.com/watch?v=4PKy3xlr… #SraVaani #ProjectVaani #SpeechAI #IndicLanguages #AIForEveryone

Celebrating the launch of SraVaani, and the conversations it has sparked. On 15 August, we open sourced SraVaani, a speech recognition model that understands 65 Indian languages and dialects.

We built SraVaani at SPIRE Lab IIsc, with @ARTPARK and supported by @Google, as an open multilingual speech model trained on spontaneous, real-world speech from Project Vaani and other public Indian corpora.

This Independence Day, we are proud to introduce SraVaani An open speech recognition model supporting 60+ Indian languages and dialects, including 40+ that current systems do not.

For Project Vaani, engagements like these bring together research, datasets and the wider speech AI ecosystem; moving us closer to our mission of enabling technology that understands and represents every Indian voice.

This is where the journey began. Watch the video to discover how a shared vision is helping ensure that every voice, every language, and every community can help shape the future of AI: youtube.com/watch?v=1cgXnSiU…

Our paper, "TVP-UNET: Threshold Variance Penalty U-Net for Voice Activity Detection in Dysarthric Speech," was presented at ICASSP 2026. Special mention to Suhas B N for representing the lab and presenting the work at the conference. #ICASSP2026 #SpeechProcessing #Dysarthria