Skip to content
View Ranjit246's full-sized avatar
🎯
Enjoying the Journey.
🎯
Enjoying the Journey.
7 followers · 35 following

Block or report Ranjit246

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Ranjit246/README.md

Speech AI Researcher Open to Collaborate

Hi, I'm Ranjit

Speech AI Researcher at Arrowhead Voice AI, Bengaluru (previously AI Researcher at Puch AI / TurboML). I build real-time conversational speech systems for Indic languages — with a focus on BFSI telephony (Hindi/Hinglish, multilingual Indic). BS-MS dual degree from IISER Berhampur (Math major, CS minor), MS thesis at IISc Bangalore.

What I'm working on

  • 🔬 Evauating Duplex Models and working on models for Indic language
  • 🏗️ Builidng Speech models (ASR, TTS) and LLMs/SLMs for Indic language solving various usecases
  • ⚡ Real-time speech-to-speech pipeline

Research interests

Indic/multilingual LM based ASR · low-latency codec TTS · edge AI for real-time translation · duplex speech mdoels · speech evaluation metrics · speech models low-resource languages

Reach me

📧 ranjitpatro200@gmail.com 💼 linkedin.com/in/ranjitpatro

Pinned Loading

  1. waveform vs Linear Spectrogram vs Me... waveform vs Linear Spectrogram vs Mel Spectrogram
    1 
    import torch, torchaudio, matplotlib.pyplot as plt
    2 
    import torchaudio.transforms as T
    3 
    
              
    
              
    
                4
                
    wav, sr = torchaudio.load("/content/7fb46edd004140c9a047aa674f7ca624.wav")
    
              
    
              
    
                5
                
    wav = wav.squeeze()
    
              
    
          
    
    
    
  
    

    2. 

      

  2. 
  
    
    
    
      
    
        
        Source Filter Separation Plot  Source Filter Separation Plot      
    

          
    
              
    
                1
                
    import torch, torchaudio, numpy as np, matplotlib.pyplot as plt
    
              
    
              
    
                2
                
    
              
    
              
    
                3
                
    wav, sr = torchaudio.load("/content/7fb46edd004140c9a047aa674f7ca624.wav")
    
              
    
              
    
                4
                
    wav = wav.squeeze()
    
              
    
              
    
                5
                
    
              
    
          
    
    
    
  
    

    3. 

      
  3. 
  
    
    
    
      
    
        
    
            
          duplex-model-exp  duplex-model-exp          Public
        
    
      
    


      
    
        
      
    

      
    
          
  
  Python


      
    
    
    
  
    

    4. 

      
  4. 
  
    
    
    
      
    
        
    
            
          Realtime-Speech-to-Speech  Realtime-Speech-to-Speech          Public
        
    
      
    


      
    
        Realtime STT, TTS and STS Pipeline with application like realtime youtube video translation with streaming
      
    

      
    
          
  
  Python


      
    
    
    
  
    

    5.