In conjunction with the podcast, we launch a series of challenges inspired by the novel ‘Ready Player One’ by Ernest Cline. We hope it will be serious fun. You can send your answers to podcast@computationalaudiology.com. The successful participants of the competition will get listed on the scoreboard above. More challenges will follow. The winner receives eternal fame and a prize awarded by the Computational Audiology Network.
Episode 1 — the intro sound–
For all easter eggs hunters (aka Gunters), we included a number of auditory phenomena and symbols hidden in the intro (our soundbite) of our podcast episode about Bayesian Active Learning. Please listen carefully and try to find all easter eggs (Computational Audiology TV, Spotify, iTunes). One symbolic gesture we will reveal. The soundbite is based on an Easter(n) European News Intro that fell into oblivion. But which? Current world affairs have shown the importance of independent news reporting and free speech. Further clues for easter eggs may be hidden on the Computational Audiology Network website (https://computationalaudiology.com/) or on Computational Audiology Network TV.
Episode 2 — automated speech recognition for the deaf–
For this challenge, you need to pass the adopted ASR (audio) Turing test explained below. Here are the rules of the game:
You need at least two players (Player One and Player two). You start in a virtual meeting (e.g. Google Meet) with ASR enabled. You are able to communicate by reading the transcription. In turns, you will mute your audio so that you can only rely on the transcript to understand the other persons in the meeting. You need to also switch off your videofeed so that others cannot see if you are reading the transcript or listening to the conversation. First Player One is randomly assigned the role audio on / audio muted by throwing a coin. Then for 10 minutes Player Two has the task to find out if Player One is listening to the conversation or reading the transcript by asking questions that Player One needs to answer. After 10 minutes Player two needs to guess if player One was using audio or only relying on the transcript. Then the players switch roles. If the players are not able to guess the audio status of the other player the technology has passed the adopted ASR Turing test. Please make a recording of the session and share your experience with us by sending a link to podcast@computationalaudiology.com. Have fun and good luck!
Classic Turing test
Imagine being in a soundproof room and in front of you is a computer terminal. You can type in questions and there will reply answers. It’s your task to determine if the responses are generated by a human or computer. If you cannot properly identify the computer it passes the Turing test.
Adopted ASR (audio) Turing test
Imagine being in a soundproof room and in front of you is a computer terminal. You can talk and there will appear replies via text (or you can hear a voice?). It’s your task to determine if the responses are generated by a human normal listener, a deaf person (aided with technology), or a computer. If you cannot properly identify the computer or the deaf person it passes the ASR (audio) Turing test.
Adopted ASR/Deepfake (audiovideo)
Turing test Imagine being in a soundproof room and in front of you is a computer terminal+speakers. You can talk and there will appear a video feed with audible responses. It’s your task to determine if the avatar/human responses are generated by a human normal listener, a deaf person (aided with technology), a deaf person signing (aided with technology to feed an avatar for text to speech) or a computer. If you cannot properly identify the computer or the deaf person it passes the ASR /Deepfake (audiovideo) Turing test.