Spotify has been granted a patent for brand new tech that may counsel songs to customers based mostly on the evaluation of their voice knowledge. The worldwide audio and music streaming large had first utilized for this patent in February 2018 and it was granted by thr United States Patent and Trademark Workplace (USPTO) this month. The patent outlines tech that may be capable to establish numerous private traits like “emotional state” of the customers and counsel content material based mostly on this mannequin. Spotify was earlier granted one other quite unsettling patent that concerned curating a person’s Spotify expertise based mostly on their persona traits.
The patent titled “Identification of style attributes from an audio sign,” describes a technique for processing an audio sign enter from customers – together with speech content material and background noise – to establish playable content material based mostly on the enter.
Within the patent description, the Swedish audio and music streaming large explains that “it is not uncommon for a media streaming utility to incorporate options that present personalised media suggestions to a person”. A typical strategy that Spotify employs to establish a listener’s music style is by utilizing fundamental info like age or gender.
The brand new patent takes it a step additional and makes use of voice recognition expertise to “acquire the style attributes of a person”. The patented expertise goals to detect the “intonation, stress, rhythm, and the likes of models of speech” and mix it with “acoustic info inside a hidden Markov mannequin structure” in order that Spotify with be capable to categorise the person’s temper as “blissful, indignant, unhappy, or impartial”.
Spotify particulars that the emotional state, gender, age, and accent of the speaker offering the voice is analysed. The metadata collected by the expertise can even point out the variety of folks within the atmosphere during which the audio sign is enter.
All this metadata is gathered and analysed together with a Spotify listener’s “earlier requests,” “listening and ranking historical past,” “hyperlinks to related profiles akin to these of the person’s mates or colleagues”, and “the person’s present music assortment and/or library”.
In line with the Spotify patent submitting, the output would be the audio output of music equivalent to the preferences, or a show of really useful subsequent music tracks equivalent to the preferences. The output may be to easily present a suggestion on a visible show.
Spotify has earlier patented one other expertise that may advocate music to customers based mostly on their private traits. This patent aimed to classify customers underneath completely different persona fashions like openness, agreeableness, extroversion, neuroticism, and conscientiousness. This sort of a categorisation had raised some eyebrows due to the seemingly unsettling persona categorisation technique utilized by Spotify.
The audio and music streaming large is understood for its algorithm that implies new audio content material for its customers. Whereas the brand new Spotify patent would not come as a shock, the effectiveness or the problematic elements that the brand new expertise entails will solely be recognized to the customers if it hits the market.
Readers ought to notice that firms patent applied sciences they’ve conceptualised, however could not essentially use the idea in a marketed product. Thus, we urge readers to take the most recent Spotify patent and course it factors the corporate as solely a chance for the long run.
Is LG Wing’s distinctive design alone sufficient to assist it achieve India? We mentioned this on Orbital, our weekly expertise podcast, which you’ll subscribe to through Apple Podcasts, Google Podcasts, or RSS, obtain the episode, or simply hit the play button beneath.