Tools Development and Model Training Service

Speechocean provides immense value to clients through technical support services in data processing tools development, model training and optimization, evaluation and Test of Speech technology.

Speechocean has an experienced IT team and develops applications and tools for data collecting, data labeling, data management, and data quality checking.  Applications offered are:

         1. Multi-channel desktop recording system

         2. Multi-channel in-car recording system

         3. Multi-channel mobile recording system

         4. Audio data transcribing tools

         5. Webpage data labeling tools Image annotation tools Speech post-processing tools: force alignment, segmentation, and feature extraction Text analyzing tools: Grapheme to Phoneme Conversion, N-gram model extraction

         ……

At the request of clients Speechocean can also provide special services in speech technology research and development, such as:

         G2P model building 
        
         Language model building

         Speech model training in languages with a huge volume of data