Compare Dragon Legal vs. Whisper

Dragon Legal is a specialized speech recognition tool designed specifically for those in the legal field, boasting a legal-centric language model crafted from an extensive database of over 400 million words derived from legal texts. This advanced software allows lawyers and legal experts to dictate documents such as contracts, briefs, and citations with impressive accuracy levels reaching up to 99%, and at a speed that is three times quicker than traditional typing methods. Users can also create personalized voice commands to streamline repetitive tasks and benefit from the ability to transcribe previously recorded audio, significantly boosting overall workflow efficiency. Dragon Legal v16 is optimized for Windows 11 and remains compatible with Windows 10, while also offering features that enhance accessibility, including the ability to playback dictated text and utilize advanced macro commands for professionals who may face physical or cognitive challenges. Furthermore, it seamlessly integrates with Dragon Anywhere Mobile, a cloud-based dictation service for both iOS and Android devices, allowing legal practitioners to maintain their productivity even while on the move. This combination of features ensures that legal professionals can work more effectively in their demanding environments.

We have developed and are releasing an open-source neural network named Whisper, which achieves levels of accuracy and resilience in English speech recognition that are comparable to human performance. This automatic speech recognition (ASR) system is trained on an extensive dataset comprising 680,000 hours of multilingual and multitask supervised information gathered from online sources. Our research demonstrates that leveraging such a comprehensive and varied dataset significantly enhances the system's capability to handle different accents, ambient noise, and specialized terminology. Additionally, Whisper facilitates transcription across various languages and provides translation into English from those languages. We are making available both the models and the inference code to support the development of practical applications and to encourage further exploration in the field of robust speech processing. The architecture of Whisper follows a straightforward end-to-end design, utilizing an encoder-decoder Transformer framework. The process begins with dividing the input audio into 30-second segments, which are then transformed into log-Mel spectrograms before being input into the encoder. By making this technology accessible, we aim to foster innovation in speech recognition technologies.