Speech AI researcher

DESCRIPTION
Role summary
We at SPIRE Lab, IISc, are seeking a Research Intern for 2 months with a possibility to convert into a Speech AI Research Associate to conduct applied and foundational research in signal processing and machine learning, with a particular focus on speech-related AI models. The role will contribute to the development, evaluation, and publication of innovative methods for speech and audio understanding, enhancement, separation, generation, coding, and real-time deployment.
The successful candidate will work collaboratively with researchers and engineers to translate research ideas into reproducible prototypes, experimental results, and technical publications.
Key responsibilities
Research and develop machine-learning models for speech and audio applications, such as speech enhancement, source separation, noise reduction, acoustic scene analysis, speaker modeling, speech coding, or speech generation.
Apply signal-processing principles to design, analyze, and improve speech and audio algorithms.
Build, train, fine-tune, and evaluate deep-learning and foundation-model approaches using speech/audio datasets.
Design rigorous experiments, establish evaluation pipelines, and communicate findings through reports, publications, presentations, and demonstrations.
Implement prototypes and collaborate on efficient or real-time deployment where relevant.
Contribute to research proposals, technical documentation, open-source codebases, and intellectual-property activities as appropriate.
Collaborate effectively across multidisciplinary teams and mentor junior researchers or students where needed.
Mandatory qualifications and technical skills
Bachelor’s degree or above in Electrical/Electronics Engineering, Computer Science, Artificial Intelligence, or a closely related discipline.
Strong foundation in digital signal processing, including time-frequency analysis, filtering, spectral methods, feature extraction, and statistical signal processing.
Demonstrated experience in machine learning and deep learning for speech, audio, or other sequential signal modalities.
Proficiency in Python and practical experience with at least one modern ML framework, preferably PyTorch; TensorFlow or JAX experience is also relevant.
Familiarity with model training, data preprocessing, experiment tracking, error analysis, and objective evaluation methods.
Understanding of core speech/audio ML tasks such as automatic speech recognition, speech enhancement, speaker separation, speech synthesis, audio classification, or audio coding.
Ability to read research literature critically and independently develop, validate, and document research ideas.
Decent written and verbal communication skills in English, including the ability to explain technical work clearly, write research reports, and collaborate constructively in a team.
Working in Linux-based environments
Nice-to-have technical skills
Experience with transformers, self-supervised learning, large speech/audio foundation models, GANs, diffusion models, or neural codecs.
Knowledge of speech and audio toolkits such as torchaudio, librosa, Kaldi, ESPnet, SpeechBrain, Hugging Face, or similar frameworks.
Experience with acoustic modeling, acoustic-articulatory modeling, spatial audio, binaural processing, beamforming, microphone arrays, or hearing-device algorithms.
Familiarity with speech datasets, data collection, annotation, augmentation, quality control, and responsible dataset use.
Docker, Git, and reproducible research workflows.
A record of peer-reviewed publications, conference presentations, patents, or open-source contributions in speech, audio, machine learning, or signal processing.
Experience supervising students, collaborating with external partners, or contributing to interdisciplinary research projects.
Desired personal attributes
Curious, rigorous, and comfortable working on open-ended research problems.
Able to balance scientific depth with practical experimentation and implementation.
Collaborative, organized, and receptive to feedback.
Comfortable presenting results to both technical and non-technical stakeholders.
Selection Procedure
2 month paid (15000 INR per month) research Internship with potential to become Research Associate based on performance
Compensation
Compensation shall be determined by qualification, experience and internship performance
Interested candidates are encouraged to apply here: https://forms.gle/ysFsR6xYHkaCzQin7
We encourage enthusiastic candidates with Engineering and management background to apply.
#Hiring #AI #IISc #BangaloreJobs #NLP #ASR #Jobs #Recruitment #urgenthiring #softwaredevelopment #voiceAI

Work arrangement
No