Speech Recognition Jobs

I have a steady stream of everyday French dialogues that must be segmented sentence by sentence, transcribed verbatim, and annotated in line with the detailed guidelines I’ll share. Each file is a two-speaker casual conversation; every utterance needs the correct O1 or O2 label, and every instance of background noise, laughter, overlap, or unclear speech must appear in the transcript using the prescribed tag format—nothing is to be omitted. The quality bar is high (98 %+ accuracy), so I’m counting on sharp listening skills, careful self-review, and a solid grasp of written French. Experience working on ASR corpora or with tools such as ELAN, Transcriber, or any similar platform is a plus, though quick learners are also welcome. Deliverables: • Sentence-level transc...

Please Sign Up or Login to see details.

I need a standalone Windows application that bundles a local GPT-style language model with a voice layer powered by Google Assistant. The workflow is simple: I open any PDF, the model ingests it locally, and then—through voice alone—I can ask questions about the document and get spoken answers back. Key points I must see working: • Runs entirely on Windows without needing command-line setup. • Uses the Google Assistant voice interface for both input and spoken output. • Processes PDFs offline: the text is extracted, fed to the model, and retained only on my machine. • Voice queries such as “What is the author’s main argument on page 5?” return an accurate, concise spoken answer within a reasonable response time. I’m open to wh...

I’m collecting clean, native-sounding Polish speech to expand a speech-recognition engine. The task is extremely simple: log in to the Mspeech web recorder, read 20 short sentences that I will supply, and submit the session. Key points • Target speakers: native Polish only, using a Standard Polish accent. • Recording setting: quiet environment—no TV, music, or street noise in the background. • Platform: Mspeech works in any modern browser on a laptop or phone; a built-in mic is fine if the room is silent. Deliverable 1. One completed Mspeech session containing all 20 sentences. Acceptance criteria – Every sentence is read exactly as written, with clear articulation. – Audio is free of background noise, clipping, or echo. –...

I need a speech-driven virtual assistant that lives on iOS, Android, and in a lightweight web client. Its core jobs are threefold: handle calendar scheduling with smart reminders, fetch facts or real-time information on request, and trigger common home-automation actions such as lights, thermostat, or security scenes. I am looking for an end-to-end build that covers natural-language understanding, speech-to-text, text-to-speech, and a small but polished UI shell for each platform. You are free to choose the NLP stack—Dialogflow, Rasa, or a custom transformer model—so long as latency stays low and the solution can scale. For voice, feel free to tap into native engines (Apple Speech, Google Speech, Web Speech) or a cloud provider like Azure Cognitive Services if it speeds deve...

I have a set of Amharic lecture recordings that need to be transcribed and lightly edited for clarity. Filler words, repeated phrases, and obvious false starts can be removed, but everything meaningful must stay intact so the final text still sounds natural and faithful to the speaker. The finished document has to be delivered as a well-formatted PDF. Please use clear paragraph breaks and consistent Amharic punctuation. A handful of specific names appear throughout the talks; I’ll send you a quick reference list before you begin. If any unfamiliar names come up, just flag them so I can confirm the spelling. No timestamps are necessary—just clean, readable Amharic text. Let me know your estimated turnaround time and any questions you might have once you review the audio.

I have several interview recordings where the participants drop their voices to a near-whisper right at the moments that matter most. I need those low-level passages isolated, enhanced and then transcribed accurately into clear, well-formatted text. Here is what the job involves: • Isolate and denoise the whispered portions without introducing artifacts or making the speech sound unnatural. • Transcribe the cleaned English dialogue with correct punctuation and speaker labels. • Flag any unintelligible words with time-stamps so I can review them quickly. I typically work with WAV and MP3 files; if you prefer a different format, let me know before we start so I can convert them for you. You’re free to use whichever tools you trust—iZotope RX, Adobe Audition, ...