Development of Transformer-Based Classification Model for Audio Datasets
Budget: ₹600 – ₹1,500 INR
Project Description:
We are seeking an experienced and highly skilled machine learning engineer or data scientist to develop a transformer-based model for the classification of audio datasets provided by the DCASE (Detection and Classification of Acoustic Scenes and Events) challenge 2023.
Project Goals:
1) Transformer Model Development: Design and implement a transformer-based neural network architecture suitable for processing audio data. The model should be capable of effectively capturing temporal dependencies and patterns present in audio sequences.
2) Data Preprocessing: Conduct comprehensive preprocessing of the DCASE 2023 audio datasets to ensure compatibility with the transformer model. This includes data cleaning, feature extraction, normalization, and any other necessary preprocessing steps.
3) Wav2Vec Transformer Model Integration: Utilize the Wav2Vec model from Hugging Face as the basis for the transformer architecture. Integrate this model into the classification pipeline and adapt it to suit the requirements of the DCASE-2023 audio dataset.
Budget: Negotiable, based on experience and proposed approach.
Interested candidates with a proven track record in machine learning and a passion for tackling challenging problems are encouraged to apply.
We are seeking an experienced and highly skilled machine learning engineer or data scientist to develop a transformer-based model for the classification of audio datasets provided by the DCASE (Detection and Classification of Acoustic Scenes and Events) challenge 2023.
Project Goals:
1) Transformer Model Development: Design and implement a transformer-based neural network architecture suitable for processing audio data. The model should be capable of effectively capturing temporal dependencies and patterns present in audio sequences.
2) Data Preprocessing: Conduct comprehensive preprocessing of the DCASE 2023 audio datasets to ensure compatibility with the transformer model. This includes data cleaning, feature extraction, normalization, and any other necessary preprocessing steps.
3) Wav2Vec Transformer Model Integration: Utilize the Wav2Vec model from Hugging Face as the basis for the transformer architecture. Integrate this model into the classification pipeline and adapt it to suit the requirements of the DCASE-2023 audio dataset.
Budget: Negotiable, based on experience and proposed approach.
Interested candidates with a proven track record in machine learning and a passion for tackling challenging problems are encouraged to apply.