Voice apparatus and methods

Abstract

An apparatus comprises: a processor 110, a memory 120 configured to store computer-readable instructions that, when executed by the processor, configure the apparatus to: receive vocal data 130 from a

Description

TECHNICAL FIELD The present disclosure relates to a voice apparatus and methods therefor, such as those used in speech and language therapy. In particular, the present disclosure relates to generating vocal exercises for a user. Aspects relate to an apparatus, a method for such an apparatus, computer program code, and a system comprising an apparatus, microphone and display screen. BACKGROUND Speech and language therapy is used to help improve a user’s communication skills. Speech and language therapy may be prescribed to a user having any one or more of a hearing impairment, cognitive delays, weak oral muscles, chronic hoarseness, cleft lip or cleft palate, motor planning problems, expressive and receptive language difficulties, dysphagia and articulation problems, which can be caused in any number of ways (e.g. injuries and / or genetic conditions). Typically, speech and language therapy comprises one or more exercises for the user to perform. For example, the user may be tasked with performing a vocal exercise, which aims to provide the desired rehabilitative effects, with the vocal exercises being provided in a physical form (e.g. a paper print-out). The same vocal exercises may be provided to all users and so are generic. Therefore, there exists a problem that users may not fully engage with their prescribed vocal exercises, because they are not interesting or varied enough to hold the user’s interest. Therefore, the user may find performing the vocal exercises repetitive or at an inappropriate difficulty level and, ultimately, stop engaging with the speech therapy and so fail to improve their communication ability. In recent years, speech and language therapy has started to be deployed using electronic devices so a user may use an apparatus (e.g. a personal mobile, dedicated speech and language therapy apparatus, personal computer and the like). Further, attempts have been made to invent forms of speech and language therapy in which a user interacts with speech and language therapy software. However, such advancements have not addressed the issue of user engagement because they merely relate to the digitisation of what went before (e.g. replacing a paper print-out with an electronic document such as a pdf). More particularly, in relation to “interactive” speech and language therapy, complete vocal exercises are currently selected from a stored database and, as such, a user may quickly perform all vocal exercises in the database and soon become uninterested in performing the vocal exercise, because they are repetitive (e.g. a user may eventually anticipate which vocal exercise will be provided to them, raising the question of whether the speech and language therapy has the desired effect). Further, again with reference to “interactive” speech and language therapy, many available apparatuses only recognise when a specific word has been said. However, there is no analytical breakdown for the user for how well the word was said, indicating where a user may improve. It is an aim of the present disclosure to address one or more of the disadvantages associated with the prior art. SUMMARY OF THE INVENTION In an aspect there is provided an apparatus comprising: a processor, a memory configured to store computer-readable instructions that, when executed by the processor, cause the apparatus to: receive vocal data from a user, the vocal data indicative of a vocal response to a vocal exercise presented to the user, wherein the vocal exercise is configured to prompt the user to provide the vocal response indicative of a vocal characteristic; analyse the vocal data to determine a performance level of the vocal exercise, the performance level indicative of the vocal characteristic indicated in the vocal response compared to a target vocal characteristic; generate a further vocal exercise in dependence on the determined performance level and profile information of the user, wherein the profile information comprises an indication of content of interest to the user; and output the further vocal exercise. The profile information of the user further may comprise one or more of: an indication of a user objective corresponding to the vocal exercises; and an indication of a difficulty level. The vocal exercise may be determined based on the profile information of the user. The apparatus may be configured to generate the further vocal exercise to prompt the user to provide a further vocal response indicative of: the same vocal characteristic, or a different vocal characteristic, in dependence on the determined performance level. The apparatus may be configured to: receive one or more previous vocal data from the user in response to a previous vocal exercise; analyse the one or more previous vocal data to determine a previous performance level of the previous vocal exercise; and determine the target vocal characteristic based on the previous performance level. The vocal characteristic may be one or more of: a maximum phonation, pitch range, jitter and shimmer, word intelligibility, sentence intelligibility, speech rate, a number of pauses, and an acoustic voice quality index. The apparatus may be configured to compare the performance level to an upper threshold and a lower threshold, wherein: in the case that the performance level is above the upper threshold, increase a difficulty level of the vocal exercise with respect to the vocal characteristic associated with the performance level; in the case that the performance level is below the lower threshold, decrease the difficulty level of the vocal exercise with respect to the vocal characteristic associated with the performance level; and in the case that the performance level is below the upper threshold and above the lower threshold, maintain the difficulty level of the vocal exercise with respect to the vocal characteristic associated with the performance level. The apparatus may be configured to: receive a user input selecting a vocal exercise from a plurality of presented vocal exercises, wherein each vocal exercise has a different format for prompting the vocal response; determine a selection frequency for each vocal exercise in dependence on the user input; and store the selection frequency in the profile information of the user, wherein the further vocal exercise is generated in dependence on the selection frequency. The apparatus, to generate the further vocal exercise, may be configured to: generate, using machine learning, one or more words based on the determined performance level and the profile information of the user; analyse the generated one or more words to determine a composition of the generated one or more words; and if the composition of the generated one or more words does not meet a predetermined criteria, re-generate the one or more words. The apparatus may be configured to: generate a results user interface, the results user interface comprising an indication of the performance level to the user, wherein the results user interface comprises the target vocal characteristic mapped to the vocal data to indicate the performance level. The apparatus may be configured to map the target vocal characteristic to the vocal data by a colour coding, the colour of the colour coding being determined by the value of the performance level. The apparatus may be configured to determine the profile information in dependence on one or more of: a received user input comprising the indication of content of interest; and a learned user history. In an aspect there is provided a method for an apparatus, the method comprising: receiving vocal data from a user, the vocal data indicative of a vocal response to a vocal exercise presented to the user, wherein the vocal exercise is configured to prompt the user to provide the vocal response indicative of a vocal characteristic; analysing the vocal data to determine a performance level of the vocal exercise, the performance level indicative of the vocal characteristic indicated in the vocal response compared to a target vocal characteristic; generating a further vocal exercise in dependence on the determined performance level and profile information of the user, wherein the profile information comprises an indication of content of interest to the user; and outputting the further vocal exercise. The method may comprise determining the vocal exercise based on the profile information of the user. The method may comprise generating the further vocal exercise to prompt the user to provide a further vocal response indicative of: the same vocal characteristic, or a different vocal characteristic, in dependence on the determined performance level. The method may comprise: receiving one or more previous vocal data from the user in response to a previous vocal exercise; analysing the one or more previous vocal data to determine a previous performance level of the previous vocal exercise; and determining the target vocal characteristic based on the previous performance level. The method may comprise comparing each of the performance level to an upper threshold and a lower threshold, wherein: in the case that the performance level is above the upper threshold, increasing a difficulty level with respect to the vocal characteristic associated with the performance level; in the case that the performance level is below the lower threshold, decreasing the difficulty level with respect to the vocal characteristic associated with the performance level; and in the case that the performance level is below the upper threshold and above the lower threshold, maintaining the difficulty level of the vocal exercise with respect to the vocal characteristic associated with the performance level. The method may comprise: receiving a user input selecting a vocal exercise from a plurality of vocal exercises, wherein each vocal exercise has a different format for prompting the vocal response; determining a selection frequency for each vocal exercise in dependence on the user input; and storing the selection frequency in the profile information of the user, wherein the further vocal exercise is generated in dependence on the selection frequency. The method, when generating the further vocal exercise, may comprise: generating, using machine learning, one or more words based on the profile information of the user; analysing the generated one or more words to determine a composition of the generated one or more words; and if the composition of the generated one or more words does not meet a predetermined criteria, re-generating the one or more words. The method may comprise: generating a results user interface, the results user interface comprising an indication of the performance level to the user, wherein the results user interface comprises the target vocal characteristic mapped to the vocal data to indicate the performance level. The method may comprise: mapping the target vocal characteristic to the vocal data by a colour coding, the colour of the colour coding being determined by the value of the performance level. The method may comprise: determining the profile information in dependence on one or more of: a received user input comprising the indication of content of interest; and a learned user history. In an aspect there is provided computer program code which, when processed by a computer, configures the computer to perform any method as disclosed herein. In an aspect there is provided a system comprising: any apparatus as disclosed herein; a microphone configured to receive the vocal response; and a display screen configured to output the further vocal exercise. Within the scope of this application it is expressly intended that the various aspects, embodiments, examples and alternatives set out in the preceding paragraphs, in the claims and / or in the following description and drawings, and in particular the individual features thereof, may be taken independently or in any combination. That is, all embodiments and / or features of any embodiment can be combined in any way and / or combination, unless such features are incompatible. The applicant reserves the right to change any originally filed claim or file any new claim accordingly, including the right to amend any originally filed claim to depend from and / or incorporate any feature of any other claim although not originally claimed in that manner. BRIEF DESCRIPTION OF THE DRAWINGS One or more examples will now be described, by way of example only, with reference to the accompanying drawings, in which: Figure 1 shows an apparatus according to examples disclosed herein; Figure 2 shows operation of an apparatus according to examples disclosed herein; Figure 3 shows operation of an apparatus according to examples disclosed herein; Figure 4 shows operation of an apparatus according to examples disclosed herein; Figures 5a-b show examples of a results interface generated by an apparatus according to examples disclosed herein; Figure 6 shows a method according to examples disclosed herein; and Figure 7 shows an example system according to examples disclosed herein. DETAILED DESCRIPTION Figure 1 shows an apparatus 100. The apparatus 100 may be suitable for providing speech and language therapy, for example, through a series of vocal exercises. The apparatus 100 may be part of a system for providing speech and language therapy. In some examples, the apparatus 100 may be a portable electronic device. A system according to examples disclosed herein is discussed in more detail with reference to Figure 5. The apparatus 100 comprises a processor 110 and a memory 120. The processor 110 may be one or more electronic processing devices 110 which operably execute computer-readable instructions. The memory 120 may be one or more memory devices. The memory 120 is electrically coupled to the processor 110. The memory 120 is configured to store computer-readable instructions, and the processor 110 is configured to access the memory 120 and execute the instructions stored thereon. That is, the memory 120 is configured to store computer-readable instructions that, when executed by the processor 110, configure the apparatus 100 to perform actions as disclosed herein. In some examples, the apparatus 100 may also comprise output means. The output means may comprise an output device configured to output one or more of visual, audio, and haptic indications. It may comprise, for example, a display screen, a touch-sensitive display screen, a speaker, and / or a haptic output device such as a vibrator. In other examples, the apparatus 100 may be functionally connected via wired or wireless connection to an external device configured to output one or more of visual, audio, and haptic indications. In this way, the output means may be configured to output the further vocal exercise. Also illustrated in Figure 1 is the operation of an apparatus 100 according to examples disclosed herein. The apparatus 100 is configured to receive vocal data 130. In some examples, the vocal data 130 may be a sound data file (e.g. an MP3 file or the like). The vocal data 130 may be received over a wired or wireless connection with a vocal data 130 collecting device, external to the apparatus 100. For example, the user may provide a vocal response into the vocal data collecting device and, in turn, the vocal data collecting device may convert the vocal response into the vocal data 130. The vocal data collecting device may be an electronic device having a microphone. In other examples, the vocal data 130 may be received directly from a user to the apparatus 100. For example, the apparatus 100 may comprise a microphone configured to receive the vocal response from a user. In this case, the apparatus 100 may be configured to convert the vocal response into the vocal data 130. The vocal data 130 is indicative of a vocal response to a vocal exercise presented to a user. The vocal exercise is configured to prompt the user to provide a response indicative of a vocal characteristic. That is, the vocal exercise may be an exercise for the user to produce a sound with the aim of rehabilitating the user’s language, voice, speech, and / or even cognition. In some examples, the vocal exercise may be used for assessing a user’s ability to produce a sound and, therefore, assess a user’s voice ability. For example, the vocal exercise may be configured to determine how long the user can sustain saying a vowel noise aloud. In other examples, the vocal exercise may be used for assessing a user’s ability to articulate sounds and words and, therefore, assess a user’s speech ability. For example, the vocal exercise may be configured to determine how intelligible a user is when speaking words and / or sentences aloud. In this way, the vocal data 130 may be indicative of voice data and / or speech data. In addition, the vocal exercise may be used for assessing a user’s ability to understand a language (i.e. a user’s cognition) and, therefore, the vocal data 130 may be indicative of language data. For example, the vocal exercise may be configured to determine how well a user describes a scene that has been displayed. Further, while the discussions herein are in relation to vocal data 130 being indicative of a vocal response (i.e. an audio input) received from a user, in some examples the data received by a user may be of another form. For example, an input type may be a touch input in response to a vocal exercise that targets the language and / or cognitive abilities of the user. For example, the vocal exercise may be a point and click exercise (e.g. asking a user to point to a named object within a displayed scene). Once the vocal data 130 has been received, the apparatus 100 is configured to analyse the vocal data 130 to determine a performance level 131 of the vocal exercise. The performance level 131 is indicative of the vocal characteristic indicated in the vocal response (i.e. the user’s attempt at the vocal exercise) compared to a target vocal characteristic. In some examples, to analyse the vocal data 130, the apparatus 100 may be configured to determine one or more quantifiable metrics associated with vocal characteristics indicated in the vocal response (i.e. the vocal data 130). The target vocal characteristic may be understood to be one or more quantifiable metrics of the vocal characteristic, prompted by the vocal exercise, having a specific value. For example, the vocal characteristic may be a speech rate and the target speech rate may be 100 words per minute. As such, the received voice data 130 may be analysed to determine an actual speech rate of a user, which can be compared to the target speech rate. In this way, the performance level 131 provides an indication for a user’s proficiency at performing the vocal exercises and producing the vocal characteristic. The performance level may be determined as an absolute difference between the vocal characteristic of the vocal response and the target vocal characteristic in some examples. The performance level may be determined as a percentage difference between the vocal characteristic of the vocal response and the target vocal characteristic in some examples. The apparatus 100 is then configured to generate a further vocal exercise 132 in dependence on the determined performance level 131 and profile information of the user. The profile information comprises an indication of content of interest to the user. That is, the apparatus 100 accounts for a user’s proficiency at a specific vocal characteristic in combination with content that is of interest to the user (e.g. has been stored as a topic of interest following a previous user questionnaire or in dependence on a user’s reaction to the subject being presented previously), when generating the vocal exercise 132. As an example, the apparatus 100 may determine that a user’s word intelligibility performance level is low and that a content of interest to the user is “football”. As such, the apparatus may be configured to generate a further vocal exercise 132 that is related to the topic of football and specifically aims to exercise a user’s word intelligibility skills, at a difficulty level associated with the user’s determined performance level 131. The apparatus 100 is configured to output the further vocal exercise 132 to the user. In some examples, the apparatus 100 may comprise output means which may be used to output the further vocal exercise. For example, the apparatus 100 may be configured to generate a vocal exercise user interface on a display screen of the apparatus 100, with the further vocal exercise 132 displayed on the display screen for the user. In some examples, the generated vocal exercise user interface may be interactive and may be configured to receive a user response (e.g. via a touch input, audio input, video input, or a combination thereof) to the generated further vocal exercise 132. In the same way, the apparatus 100 may be configured to output the initial vocal exercise using the output means and a vocal exercise user interface. Figure 2 illustrates operation of an apparatus 100 according to examples disclosed herein. As shown in Figure 2, the apparatus is configured to receive the voice data 130. The voice data 130 may be indicative of a user input 208 (e.g. vocal input (a vocal response), a touch input, a video input or the like) to a vocal exercise presented to the user. The user input 208 is received by a training unit 210 in functional connection with the apparatus 100, either externally or internally (e.g. the processor 110) to the apparatus 100. The training unit 210 may be configured as a vocal data collecting device, which may convert the user input 208 into the vocal data 130. That is, the training unit 210 may enable a user to perform the vocal exercises and record their response 208. A processing unit212 (e.g. processor 110 of the apparatus 100) may be configured to analyse the vocal data 130 to determine a performance level 131 of the vocal exercise. The performance level 131 may be indicative of the vocal characteristic indicated in the vocal response 208 compared to a target vocal characteristic. The performance level 131 is fed back to other processing steps in Figure 2 as discussed below. The vocal characteristic may be one or more of a maximum phonation, pitch range, jitter and shimmer, word intelligibility, sentence intelligibility, speech rate, a number of pauses, and an acoustic voice quality index. The vocal characteristic may be one or more of a naming of a picture, a scene selection, a scene description, word finding, reversible passive sentences, story and image memory, word category, position relationship, and a token test. Maximum phonation may be understood as how long a user can sustain making a specific noise aloud. For example, how long a user can sustain saying a vowel sound aloud. In this case, the voice data may be a sound file of the user saying “ / Ah / ” or " / Oh / ", for example, for as long as possible. Pitch range may be understood as how much a user can vary the pitch of their voice. For example, varying the pitch of the user’s voice from a starting pitch to a lower pitch, and / or varying the pitch of the user’s voice from a starting pitch to a higher pitch. Jitter and shimmer may be understood to relate to a variation in frequency (jitter) and amplitude (shimmer) of a user’s voice when uttering the same (or different) vowel for a long time. Word intelligibility may be understood as how well a user can speak a word and their phonemes. For example, how completely and accurately a user can say a word. Sentence intelligibility may be understood as how well a user can speak one or more sentences. For example, how completely and accurately a user can say one or more sentences. Speech rate may be understood as how fast a user can speak one or more known sentences. A number of pauses may be understood as the number of pauses (e.g. no phonation I sound) a user takes in the total time duration of providing the vocal response. Acoustic voice quality index may be understood as a measure of voice quality. Naming of a picture may be understood as determining whether a user can describe an image using one or more words. A scene selection may be understood as determining whether a user can select a named item within an image containing several items (i.e. a scene). A scene description may be understood as determining whether a user can answer questions regarding an item in a scene. Word finding may be understood as determining how many words in a certain category (e.g. animals, fruits etc) the user can say. Reversible passive sentence exercises may be understood as determining whether the user can understand, construct and / or speak sentences in the reversible passive form (e.g. when presented with the sentence “The detectives were tricked by the computer hacker” the user’s task is to speak the sentence “The computer hacker was tricked by the detectives”). Story and image memory may be understood as determining whether a user can comprehend and remember details in a story or image and answer questions about the story. Word category may be understood as determining if the user can identify whether images and / or words belong to certain categories (e.g. syntactic class (e.g. nouns, verb, etc), or semantic class (e.g. animals, fruits, etc)). Position relationship may be understood as determining whether the user can identify a relative position between objects in a scene. A token test may be understood as determining if the user can identify a quality of a presented image, such as shapes, colours, and sizes. Maximum phonation, pitch range, and jitter and shimmer may be used to assess a user’s ability to use their voice. Therefore, vocal data 130 associated with maximum phonation, pitch range, and jitter and shimmer may be understood as examples of voice data. Further, word intelligibility, sentence intelligibility, speech rate and a number of pauses may be used to assess the user’s ability to speak. Therefore, vocal data associated with word intelligibility, sentence intelligibility, speech rate and a number of pauses may be understood as examples of speech data. Further, vocal data associated with naming of a picture, a scene selection, a scene description, word finding, reversible passive sentences, story and image memory, word category, position relationship, and a token test may be used to assess the user’s understanding and proficiency of a language and, therefore, understood as language data. Each vocal characteristic is different in terms of what is required in a vocal response from a user and, as such, may have one or more different quantifiable metrics associated therewith. Therefore, analysis performed by the processor 110 to determine the performance level 131 of the vocal exercise may depend on the vocal characteristic being assessed, and therefore on the vocal exercise being performed. Thus, different vocal characteristics may have different quantifiable metrics. In the example of a maximum phonation, the quantifiable metric may comprise one or more of a longest phonation duration; a number of phonation segments and a number of pauses; a duration of each phonation and pause; and a total phonation duration (e.g. the sum of all phonation segments). In the example of a pitch range, the quantifiable metric may comprise one or more of a minimum, maximum, range, standard deviation and a mode of the pitch of the vocal response. In the example of a jitter and shimmer, the quantifiable metric may comprise one or more of a relative and absolute (in seconds and decibels (dB)) jitter (i.e. frequency) and shimmer (i.e. amplitude) of the vocal response. In the example of word intelligibility and sentence intelligibility, the quantifiable metric may be one or more of a completeness, accuracy, and fluency of the whole word, an accuracy score for each of the phonemes of the target word; target word phonetic characteristics (e.g. number of syllables, presence of diphthongs, affricative, and the like); and accuracy by phoneme characteristics in the vocal response. In the example of speech rate, the quantifiable metrics may comprise one or more of a number of syllables spoken over a total time duration; a number of syllables spoken over a total phonation duration; and an average syllable duration in the vocal response. In the example of a number of pauses, the quantifiable metrics may comprise one or more of a total number of pauses over a time duration; the duration of any pauses; and how frequently the pauses occur in the vocal response. In the example of an acoustic quality index, the quantifiable metrics may comprise one or more of a smoothed cepstral peak prominence of the voice data, a harmonics-to-noise ratio of the voice data, a local shimmer of the voice data, a gradient of a long-term average spectrum of the voice data; and a tilt of a trendline through the long-term average spectrum of the voice data. The performance level 131 is indicative of the vocal characteristic indicated in the user’s vocal response compared to a target vocal characteristic. The target vocal characteristic may be understood to be a quantifiable metric of the vocal characteristic having a specific value. For example, if the vocal exercise is a pitch range exercise, the target vocal characteristic may be a specific range of frequencies of a user’s vocal data 130 to be achieved. In this way, the target vocal characteristic may be understood as a vocal characteristic aim to be achieved by a user. The target vocal characteristic may be respective for each user of the apparatus 100. For example, the target vocal characteristic may be stored in profile information of each user. The target vocal characteristic may be determined by a qualified professional (e.g. a speech and language therapist or the like) through a series of one or more assessments. For example, a user may visit the qualified professional, perform the one or more assessments, the results of which will be determined by the qualified professional, who will identify the target vocal characteristic, to help the user improve their communication and / or language skills. In some examples, the apparatus 100 may determine the target vocal characteristic in isolation or in addition to a target vocal characteristic determined by a qualified professional. The apparatus 100 may be configured to receive one or more previous vocal data 130 from the user in response to a previous vocal exercise. For example, the previous vocal data 130 may be stored in the memory 120 of the apparatus 100, in a cloud storage facility, or in any device external to the apparatus 100 that is in wired or wireless communication with the apparatus 100. The apparatus 100 may analyse the one or more previous vocal data 130 to determine a previous performance level 131 of the previous vocal exercise. The apparatus 100 may determine the target vocal characteristic based on the previous performance level 131. For example, by the apparatus 100 determining a previous performance level 131, the apparatus 100 aims to determine where the user struggles to communicate / use language effectively. In this way, the target vocal level is targeted towards a user’s needs and therefore increases the effectiveness of the vocal exercises. The performance level 131 is indicative of a vocal characteristic indicated in the vocal response compared to a target vocal characteristic. The analysis by the apparatus 100 to determine the performance level 131 may comprise determining one or more quantifiable metrics of the user’s vocal response in dependence on the vocal exercise being performed (i.e. because each vocal characteristic (and thus vocal exercise) has respective quantifiable metrics). The apparatus 100 may then compare the determined one or more quantifiable metrics with those of the target vocal characteristics. The performance level 131 may be determined as one or more of an absolute difference between the vocal characteristic of the vocal response and the target vocal characteristic; and a percentage difference between the vocal characteristic of the vocal response and the target vocal characteristic. In some examples, the performance level 131 may additionally take into account and indicate a time taken for users to generate a vocal response and / or a number of determined pauses (e.g. periods of time with no phonation I sound detected) within the vocal response. For example, a quantifiable metric of a vocal characteristic comprised in the vocal response may compare closely with that of a target vocal characteristic. However, a number of pauses present in the vocal response (and time take to deliver the vocal response) may be high. In such a case, the apparatus 100 may be configured to determine a low performance level with respect to the vocal characteristic because the high number of pauses present in the vocal response may indicate that the user does not have a high proficiency regarding the vocal characteristic. As each vocal characteristic is related to respective one or more quantifiable metrics, the apparatus 100 may store one or more performance levels 131 for each of the vocal characteristics, which may be utilised when generating the further vocal exercise 132 as shown in Figure 2. In this way, the apparatus 100 may be configured to take a user’s previous performance into account when generating the further vocal exercise 132. For example, the apparatus 100 may use the performance level to determine a difficulty level of the further vocal exercise 132. For example, if the user finds difficulty in providing a vocal response to a vocal exercise (e.g. a pitch range), the performance level 131 associated with that exercise may not be sufficiently high enough (i.e. be below a target threshold), and the further vocal exercise 132 may be set at a lower difficulty level (i.e. the further vocal exercise 132 will be of an easier difficulty) than before. In the example of the vocal exercise being a pitch range, the lower difficulty level may be represented by a narrower pitch range, with a lower decibels volume, to be achieved. Similarly, if the user proficiently provides a vocal response to a vocal exercise, the performance level 131 associated therewith may be sufficiently high, exceeding a current difficulty level, and the further vocal exercise 132 may then be set at a higher difficulty level (i.e. the further vocal exercise 132 will be of an harder difficulty) than before. In the example of the vocal exercise being a pitch range, the higher difficulty level may be represented by a wider pitch range, with a higher decibels volume, to be achieved. The apparatus 100 may be configured to generate the further vocal exercise 132 to prompt the user to provide a further vocal response indicative of: the same vocal characteristic, or a different vocal characteristic, in dependence on the determined performance level 131. For example, the apparatus 100 may determine that the determined performance level 131 indicates that the user is or is not acceptably performing the target vocal characteristics. In some examples, if the determined performance level 131 indicates that a user is not acceptably performing the target vocal characteristic, the apparatus 100 may be configured to generate the further vocal exercise 132 to prompt the user to provide a further vocal response indicative of the same vocal characteristic. For example, the apparatus 100 may determine that the performance level is not sufficiently high enough and, as such, the same vocal characteristic should be assessed in the further vocal exercise 132. In other examples, if the determined performance level 131 indicates that a user is acceptably performing the target vocal characteristic, the apparatus 100 may be configured to generate the further vocal exercise 132 to prompt the user to provide the further vocal response indicative of a different vocal characteristic. For example, the apparatus 100 may determine that the performance level is sufficiently high enough and, as such, a different vocal characteristic should be assessed in the further vocal exercise 132. The apparatus 100 may be configured to generate a further vocal exercise 132 in dependence on the determined performance level 131 and profile information of the user. As above, the performance level 131 is indicative of a vocal characteristic indicated in the vocal response compared to a target vocal characteristic. In reference to Figure 2, the determined performance level 131 may be comprised within user information 206. The user information 206 may be understood as information relating to the user. For example, the user information 206 may comprise the determined performance level 131 and the information comprised within the profile information. For example, one or more of the determined performance level 131, a user objective information, a difficulty level, and user preference information may be part of the profile information. The profile information comprises at least an indication of content of interest 204b to the user. The content of interest 204b may be understood to relate to a topic of subject matter (e.g. “football”). The apparatus 100 may be configured to determine the profile information in dependence on one or more of: a received user input 202 comprising the indication of content of interest 204b; and a learned user history. The apparatus 100 may be configured with functionality to allow the user to self-report content of interest to the apparatus 100. In some examples, the apparatus 100 may display one or more topics of subject matter, from which the user can select the content of interest. The apparatus 100 may be configured to store the selected content of interest in the memory 120. The apparatus 100 may be configured to determine the profile information comprising the indication of content of interest 204b based on the selected content of interest. The apparatus 100 may be configured to determine the profile information in dependence on a learned history. In some examples, the apparatus 100 may be configured to determine, through the use of a machine learning algorithm, the content of interest 204b. For example, the apparatus 100 may store previous user inputs to the apparatus 100 (e.g. selections of vocal exercises, voice data received during vocal exercises, user inputs during self-reporting content of interest) in the memory 120. The apparatus 100 may be configured to use the stored previous user input to the apparatus 100 as an input to the machine learning algorithm configured to determine a content of interest associated therewith. That is, the learned user history may comprise a determined content of interest 204b indicative of content that is typically engaged with by a user. The apparatus 100 may be configured to store the determined content of interest in the memory 120 as part of the profile information associated with the user. Therefore, the apparatus 100 may be configured to determine the profile information comprising a determined content of interest 204b based on the learned user history. The indication of content of interest 204b may be determined by an interest collection unit 204a. The apparatus 100 may, upon first use or any subsequent use, be configured to enable selfreporting of the indication of the content of interest. For example, a list of subjects and / or topics may be presented to a user using output means of the apparatus 100, or an output means of an external device connected to the apparatus 100. A user input may be received with respect to one or more of the subjects and / or topics in the list, thereby providing an indication of the content of interest for a user. Further, the profile information may comprise an indication of a user objective corresponding to the vocal exercises, and / or an indication of a difficulty level. In reference to Figure 2, the user objective and the indication of a difficulty level may be comprised within user information 206. The user objective may indicate what exercise type a vocal exercise is. For example, how to improve a user’s voice, speech, language and / or use of facial muscles. As such, the user objective may be understood as the overall reason why a user engages with the apparatus 100 and the user objective may represent a specific goal for the user. In this way, the further vocal exercise 132 is generated to account for a user objective and, as such, the apparatus 100 aims to provide a new interaction between the user and the apparatus 100 that is more efficient in improving the user’s vocal performance. For example, a user may spend less time interacting with the apparatus 100 as a result of more suitable vocal exercises 132 being presented to the user consistently. The user objective may be determined by the apparatus 100. The apparatus 100 may be configured to determine the user objective based on previous performance with respect to vocal exercises. More specifically, the apparatus 100 may be configured to determine the user objective based on the one or more determined performance levels 131 for previous vocal exercises. For example, the apparatus 100 may determine, based on determined performance levels 131, that a user particularly struggles with vocal exercises pertaining to speech. As such, the apparatus 100 may determine that the user objective should relate to improvement of a user’s speech. The apparatus may be configured to store the user objective in the memory 120 as part of the profile information. Therefore, when generating the further vocal exercise 132, the apparatus 100 is configured to account for the stored user objective, comprised within the profile information. In reference to Figure 2, the user objective information may be understood to be comprised within the user information 206. In some examples, the user objective may be supplemented by a qualified professional, as indicated by a qualified professional input 218 and treatment suggestion information 214 in Figure 2, determined in the same way as the target vocal characteristic. In this way, the further vocal exercise 132 can follow a qualified guidance. The indication of a difficulty level may be understood as a required proficiency at a vocal characteristic to achieve a suitably high performance level. For example, a high difficulty level for a pitch range vocal exercise may relate to a broad range of pitches with which the user can sweep while holding a vowel sound aloud at a high volume. Alternatively, a low difficulty level for a pitch range vocal exercise may relate to a narrow range of pitches with which a user can sweep while holding a vowel sound aloud at a lower volume. As another example, a high difficulty level for a speech rate exercise may relate to a high speech rate with which the user can speak, while maintaining intelligibility. Alternatively, a low difficulty level for a speech rate exercise may relate to a lower speech rate with which the user can speak, while maintain intelligibility. The difficulty level may be determined by the apparatus 100 and / or a qualified professional (e.g. comprised as part of the treatment suggestion information 214). The profile information may further comprise user preference information. The user preference information may indicate one or more of: words that the user does not want included in a vocal exercise (e.g. swearwords); a preferred maximum number of syllables for words included in a vocal exercise; a preferred minimum number of syllables for words included in a vocal exercise; and a preferred format of vocal exercise (e.g. a simple reading task, a quiz, facts, karaoke, and point and click, or the like). As such, when generating a further vocal exercise 132, such user preferences may be accounted for by the apparatus 100. In reference to Figure 2, the user preference information may be comprised within the user information 206. The apparatus 100 may be configured to determine the user preference information in dependence on one or more of a received user input 202 comprising an indication of the user preference information and a learned user history. Similarly to the content of interest 204b, the apparatus 100 may be configured to provide the user with an opportunity to self-report user preference information into the apparatus 100. As an example, the apparatus may display one or more format types of vocal exercises, from which the user can select a most preferred. The apparatus 100 may be configured to store the selected user preference information in the memory 120. In other examples, the apparatus 100 may be configured to determine the user preference information, based on a learned user history. A vocal exercise may be available in a number of different formats. For example, the vocal exercise may be in the form of a simple reading task, a quiz, facts, karaoke, and point and click, or the like. As such, a user may have a preference for a particular format that they engage with more than other formats. In some examples, the apparatus 100 may be configured to receive a user input selecting a vocal exercise. For example, the selected vocal exercise may be selected from a plurality of presented vocal exercises, wherein each vocal exercise has a different format for prompting the vocal response. The apparatus 100 may be configured to determine a selection frequency for each vocal exercise in dependence on the user input. For example, the apparatus 100 may be configured to store an indication of the number of times a vocal exercise has been selected by a user (via the user input). Therefore, each time a vocal exercise is selected by a user (via the user input), the indication of the number of times a vocal exercise has been selected by a user is increased accordingly. In this way, the apparatus 100 is able to determine which vocal exercise from the plurality of vocal exercises is selected most frequently by the user. The apparatus 100 may be configured to store the selection frequency in the profile information of the user. In this way, because the further vocal response is determined based on the performance level 131 and the profile information of the user, the further vocal response may be generated in dependence on the selection frequency, stored in the profile information. To generate the further vocal exercise 132, the apparatus 100 may be configured to generate one or more words based on the determined performance level 131 and the profile information of the user. That is, the apparatus 100 may generate the one or more words based on information that has been determined and / or received by the apparatus 100 (e.g. the indication of a content of interest of the user, user objective, difficulty information, and / or user preferences comprised within the profile information, and the determined performance level 131). In this way, the one or more words (and therefore further vocal exercise 132) may be understood to be generated by the apparatus 100 using machine learning. The one or more words may be generated from one or more sources accessible to the apparatus 100. For example the apparatus 100 may be configured to access one or more of a local storage (e.g. the memory 120), and a remote storage. For example, the apparatus 100 may be configured to access one or more of a database and a web search engine. In some examples, one or more of the local and remote storages may store within a database comprising one or more words. Each of the one or more words may have metadata associated therewith. The metadata may comprise a topic indication. The topic indication may comprise one or more descriptors indicating the subject matter that a word relates to. That is, the topic indication may be a semantic class of a word. In this way, the topic indication may be utilised by the apparatus 100 when generating the further vocal exercise 132 in dependence on the profile information comprising the indication of content of interest to the user. As a nonlimiting example, the indication of content of interest to the user may be “football”, the apparatus 100 may generate the one or more words by searching through the database and determining words, for use in the further vocal exercise 132, having metadata relating to the word “football”. In this example, words having topics such as “sport”, “active lifestyle”, “health”, or the like, may be considered usable. In some examples, the metadata may comprise one or more of, in addition to the one or more topics of an associated word, an objective indication, a vocal characteristic indication, a difficulty level indication, a syntactic class indication, and a frequency indication, each being associated with the word to which the metadata is associated. The objective indication may relate to indicate what exercise type a word may be associated with. For example, the objective indication may indicate whether a word is associated with one or more of a voice, speech, language, and a facial muscle exercise type. In this way, the objective indication may be utilised by the apparatus 100 when generating the further vocal exercise 132 in dependence on the profile information comprising the user objective corresponding to the vocal exercises. For example, the user objective may indicate a speech exercise type. As such, the apparatus 100 may be configured to generate the one or more words using words having an objective indication for speech exercise type. This may, for example, be one or more words that are particularly challenging to pronounce and thus may help exercise and assess a user’s speech ability. The vocal characteristic indication may indicate which vocal characteristic may be best exercised with the word. For example, the vocal characteristic indication may indicate whether a word is associated with one or more of a maximum phonation, pitch range, jitter and shimmer, word intelligibility, sentence intelligibility, speech rate, a number of pauses, and an acoustic voice quality index. In this way, the vocal characteristic indication may be utilised by the apparatus 100 when generating the further vocal exercise 132 in dependence on the profile information comprising the user objective corresponding to the vocal exercises. The difficulty level indication may relate to how challenging a word may be to understand and / or pronounce for a user. For example, the difficulty level indication may indicate one or more of whether a word is multisyllabic and therefore, may be difficult for a user to pronounce; and whether the word comprises phonemes that are typically difficult to pronounce (e.g. “thorough”, “shriek”, “asterisks”). In this way, the difficulty level indication may be utilised by the apparatus 100 when generating the further vocal exercise 132 in dependence on the profile information comprising the difficulty level. The syntactic class indication may relate to what type of word a word is. For example, the syntactic class may indicate whether a word is a verb, noun, adjective, connective, and an article. In this way, the apparatus 100 may be able to identify whether a user has particular strengths and / or weaknesses in relation to words belonging to certain syntactic classes. As such, the syntactic class may be utilised by the apparatus 100 when generating the further vocal exercise 132. The frequency indication may relate to how frequently a word is used in a vernacular. For example, the word “accismus" may not be spoken frequently in the vernacular of a user’s geographical region and therefore, may be important for a user to practise. Alternatively, the apparatus 100 may actively select words that are not spoken frequently in the vernacular, as a means of increasing the difficulty of a vocal exercise (e.g. to test a user’s ability to read and comprehend a word that is not frequently encountered). In this way, the frequency indication may be utilised by the apparatus 100 when generating the further vocal exercise 132. The apparatus 100 may be configured to generate the one or more words based on the determined performance level 131 and the profile information of the user. Initially, the apparatus 100 may be configured to generate the one or more words using a first method. Using the first method, the apparatus 100 may search the database to generate the one or more words, ensuring that the generated one or more words is associated with the determined performance level 131 and one or more of the components within the profile information. As an example, the profile information may comprise an indication of content of interest to the user (e.g. football), an indication of a user objective (e.g. facial muscle rehabilitation), a difficulty level, and a user preference relating to a preferred maximum number of syllables for words included in a vocal exercise. Using the first method, the apparatus 100 may generate one or more words, having, overall, metadata corresponding to the indication of the content of interest to the user and the indication of a user objective, while the generated one or more words may not strictly correspond to the remaining components of the profile information (i.e. the difficulty level and user preference relating to a preferred maximum number of syllables for words included in a vocal exercise). For example, if the user preference relating to a maximum number of syllables for words included in a vocal exercise is stored as three syllables, the apparatus 100, using the first method, may include words having four or more syllables in the generated one or more words. However, due to the statistical nature of machine learning, the apparatus 100 may be configured, using the first method, to generate one or more words that is associated with the determined performance level 131 and all of the components within the profile information. Once the one or more words have been generated, the apparatus 100 may be configured to analyse the generated one or more words. For example, the apparatus 100 may determine a composition of the generated one or more words. That is, the apparatus 100 may analyse the metadata of the one or more words to determine the overall characteristics of the one or more words. In an example, if the apparatus 100 generates one word, the overall characteristics of the one or more words will correspond to the metadata of the generated one word. In another example, if the apparatus 100 generates more than one word, the overall characteristics of the more than one words will correspond to a combination of the metadata of the generated more than one words. The apparatus 100 may be configured to compare the generated one or more words with a predetermined criteria. The predetermined criteria may be understood as a criteria for determining the suitability of the generated one or more words with respect to a user’s performance level 131 and profile information. The predetermined criteria may be a requirement that the generated one or more words having overall characteristics that match the performance level and match two or more of the components of the profile information. The predetermined criteria may be determined by the user and / or by a qualified professional. For example, a user may input the number of components of the profile information that are to be adhered to. In the case that the generated one or more words meets the predetermined criteria, the apparatus 100 may be configured to output the generated one or more words, in the form of the further vocal exercise 132, to the user. In the case that the generated one or more words does not meet the predetermined criteria, the apparatus 100 may be configured to re-generate the one or more words. In this example, the re-generated one or more words may be compared to the predetermined criteria to determine the suitability of the re-generated one or more words for the user. In some examples, the apparatus 100 may be configured to re-generate the one or more words in the same way the apparatus 100 generated the one or more words initially. That is, the apparatus 100 may re-generate the one or more words using the first method. In other examples, the apparatus 100 may be configured to re-generate the one or more words using a second method. Using the second method, the apparatus 100 may search the database to re-generate the one or more words, ensuring that the re-generated one or more words is associated with the determined performance level 131 and two or more of the components within the profile information. In this way, the second method may be understood as a more thorough, and computationally taxing, method than the first method of searching the database and generating the one or more words. Advantageously, by first searching the database and generating the one or more words using the first method and, only if the resulting one or more words do not meet the predetermined criteria, subsequently using the second method, the apparatus 100 is able to expend further computation load only when necessary. In this way, the apparatus 100 may operate in a way to improve latency and reduce computational load when generating the further vocal exercise 132. In reference to Figure 2, the determined performance level 131, the user information 206, and the treatment suggestion information 214, may be used as input into a personalisation unit 216 that may be in functional connection with the apparatus 100, either externally or internally (e.g. as part of the processor 110). In some examples, the personalisation unit 216 may generate the one or more words, based on the information input thereto. In this way, the personalisation unit 216 may be configured to create the appropriate content comprised within the further vocal exercise 132. Once the further vocal exercise has been generated, the personalisation unit 216 may be configured to generate an interactive exercise. For example, the personalisation unit 216 may generate an interactive user interface, through which a user may input a response 216 to the first vocal exercise 132. Figure 3 shows the operation of the apparatus 100, in relation to the personalisation unit 208. With reference to Figure 3, the apparatus 100 may be configured to receive the user information 206. In response to receiving the user information 206, the apparatus 100 may be configured to identify a suitable exercise type, therapy target, difficulty level of an exercise, and exercise format, illustrated at 302. For example, the user information 206 may comprise the user objective, difficulty information, and or user preferences as determined by the apparatus 100 and / or as self-reported by a user to determine the exercise type, therapy target, difficulty level, and exercise format. Then, the apparatus 100 may be configured to receive the content of interest 204b. In response, the apparatus 100 may search a database comprising one or more words to generate content for the further vocal exercise 132, illustrated at 304. For example, the apparatus 100 may search the database using the content of interest 204b to generate one or more words having a topic related to the content of interest 204b. In some examples, the content retrieved from the database may be further supplemented with the use of a web search. That is, the apparatus 100 may be configured to search for content related to the content of interest 204b via a web search engine. In some optional examples, indicated by the dashed line in Figure 3, the apparatus 100 may be configured to receive one or more treatment suggestion 214 to supplement one or more of the determined one or more words and determined exercise type, therapy target, difficulty level of an exercise, and exercise format, illustrated at 306. Based on this, the apparatus 100 may be configured to produce the finalised exercise content 308 that may be used in the further vocal exercise 132. The apparatus 100 may generate an interactive user interface, through which a user may input a response 208 to the first vocal exercise 132, illustrated at 308. In this way, the interactive user interfaces may provide a means for communicating the finalised exercise content 308, to generate the further vocal exercise 132. The above discussion relates to the further vocal exercise 132 that is generated by the apparatus 100 in dependence on the determined performance level 131 and the profile information of the user. In some examples, the vocal exercise initially displayed to the user, from which the performance level is determined, may be determined by the apparatus 100 based on the profile information of the user. In other examples, the vocal exercise may be determined by the apparatus 100 based on the profile information of the user and a most recent performance level 131 for the vocal exercise. Figure 4 illustrates operation of an apparatus 100 according to an example. More specifically, Figure 4 illustrates how a difficulty level may be determined for a user. The apparatus 100 may be configured to compare the performance level 131 to an upper threshold and a lower threshold. The upper and lower thresholds may be representative of upper and lower bounds to a difficulty level. For example, a speech rate vocal characteristic may be associated with a number of difficulty levels, wherein each difficulty level may have a minimum and maximum speech rate associated therewith. As another example, a sentence intelligibility vocal characteristic may be associated with a number of difficulty levels, wherein each difficulty level may be characterised by one or more of: a length of the vocal exercise (e.g. how many words and / or sentences are present in the vocal exercise); a percentage of words in the vocal exercise that are equal to or more than a predetermined number of syllables (e.g. three syllables), and a percentage of words in the vocal exercise that contain clusters (e.g. the presence of two or more consonants in sequence). In an example, in the case that the performance level 131 is above the upper threshold, the apparatus 100 may be configured to increase a difficulty level of the vocal exercise with respect to the vocal characteristic associated with the performance level 131. In the case that the performance level 131 is below the lower threshold, the apparatus 100 may be configured to decrease the difficulty level of the vocal exercise with respect to the vocal characteristic associated with the performance level 131. In the case that the performance level 131 is below the upper threshold and above the lower threshold, the apparatus 100 may be configured to maintain the difficulty level of the vocal exercise with respect to the vocal characteristic associated with the performance level 131. The performance level 131 is indicative of the vocal characteristic indicated in the user’s vocal response compared to a target vocal characteristic. A vocal characteristic may be indicated by one or more quantifiable metrics. Each quantifiable metric may be associated with a respective performance level 131, an upper threshold, and a lower threshold. In an example, the apparatus 100 may be configured to determine the one or more quantifiable metrics 402. The apparatus 100 may then be configured to, for each quantifiable metric compare a quantifiable metric to its respective upper threshold 404. If the quantifiable metric is determined to be more than its respective upper threshold, the apparatus 100 may be configured to increase the difficulty level of the vocal exercise with respect to the vocal characteristic 406. The apparatus 100 may be configured to store the newly increased difficulty level in the profile information for the user for subsequent use when generating further vocal exercises 132. If the quantifiable metric is determined to be below its respective upper threshold, the apparatus 100 may be configured to compare the quantifiable metric to its respective lower threshold 408. If the quantifiable metric is determined to be less than its respective lower threshold, the apparatus 100 may be configured to decrease the difficulty level of the vocal exercise with respect to the vocal characteristic 410. The apparatus 100 may be configured to store the newly decreased difficulty level in the profile information for the user for subsequent use when generating further vocal exercises 132. If the quantifiable metric is determined to be above its respective lower threshold, overall, the quantifiable metric will be determined to be both below its respective upper threshold and above the lower threshold. In this case, the apparatus 100 may be configured to maintain the difficulty level with respect to the vocal characteristic 412. Figures 5a-5b show examples of a results user interface generated by an apparatus 100 according to examples. In some examples, the apparatus 100 may be configured to generate a results interface 500a, 500b. The results user interface 500a, 500b may comprise an indication of the performance level 131 to the user. In some examples, the indication of the performance level 131 may be a numerical indication. For example, the indication of the performance level 131 may be a percentage difference between the vocal characteristic indicated in the vocal response compared to the target vocal characteristic. The results user interface may also comprise the target vocal characteristic mapped to the vocal data 130 to indicate the performance level 131. For example, the target vocal characteristic may be mapped to the vocal data 130 by a colour coding. In some examples, the colour coding may be determined by the value of the performance level 131. For example, the apparatus 100 may be configured to assign specific colours to specific values of the performance level 131. In the non-limiting example of Figures 5a-5b, the vocal characteristic may be at least word intelligibility. From Figures 5a-5b, the performance level 131 may be indicated by a percentage indication 502a, 502b. Further, the mapping of the target vocal characteristic to the vocal data 130 may be indicated by the colour in which the target vocal characteristic (i.e. the word to be pronounced by the user as a test of word intelligibility) is indicated. For example, when a user does not sufficiently pronounce a word and / or part of a word, as part of a word intelligibility vocal exercise, the apparatus 100 may be configured to represent the corresponding word and / or part of the word in the generated results user interface in a way distinguished from the words and / or parts of words that have been sufficiently pronounced. For example, in Figures 5a-5b, words and / or parts of words that are not sufficiently pronounced may be represented in a darker colour 504a, 504b than the words and / or parts of words that have been sufficiently pronounced. Other examples may not be so limited to a grayscale representation. For example, any colour may be used (e.g. red and green). In this way, the results user interface aims to facilitate easier interpretation of the performance level 131 for a user. In some examples, the apparatus 100 may be configured to generate and output the results user interface 500a, 500b for the user once the performance level 131 has been determined. In other examples, the apparatus 100 may be configured to generate and output the results user interface 500a, 500b for the user once a predetermined number of vocal exercises have been completed by the user. As seen in Figure 5a, as a schematic example, the results user interface 500a may comprise results for 8 vocal exercises. As seen in Figure 3b, as a schematic example, results user interface may comprise results for 3 vocal exercises. Figure 6 shows a method 600 according to examples disclosed herein. Figure 6 shows a method 600 for an apparatus 100. The method comprises: receiving 602 vocal data 130 from a user, the vocal data 130 indicative of a vocal response to a vocal exercise presented to the user, wherein the vocal exercise is configured to prompt the user to provide the vocal response indicative of a vocal characteristic; analysing 604 the vocal data 130 to determine a performance level 131 of the vocal exercise, the performance level 131 indicative of the vocal characteristic indicated in the vocal response compared to a target vocal characteristic; generating 606 a further vocal exercise 132 in dependence on the determined performance level 131 and profile information of the user, wherein the profile information comprises an indication of content of interest to the user; and outputting 608 the further vocal exercise 132. The method 600 may be performed by the apparatus 100 illustrated in Figure 1. In particular, the memory 120 may comprise computer-readable instructions (e.g. computer software) which, when executed by the processor 110 of the apparatus 100 disclosed herein, perform a method 600 as disclosed herein. Also disclosed herein is a computer software which, when executed by one or more electronic processors of an apparatus 100 as disclosed herein, is arranged to perform a method 600 as disclosed herein. The blocks illustrated in Figure 6 may represent steps in a method 600 and / or sections of code in a computer program configured to control the apparatus 100 as described above to perform the method steps. The illustration of a particular order to the blocks does not necessarily imply that there is a required or preferred order for the blocks and the order and arrangement of the block may be varied. Furthermore, it may be possible for some steps to be omitted or added in other examples. Figure 7 shows a system 700 according to an example. The system comprises an apparatus 100. For example, the apparatus 100 may be any apparatus 100 as disclosed herein. Further, the system 700 comprises a microphone 702; and a display screen 704. The microphone 704 is configured to receive vocal data 130 from a user. The display screen 704 is configured to output the further vocal exercise 132 and may be configured to output the vocal exercise. In some examples, the display screen 704 may be comprised within the output means of the apparatus 100. In other examples, the display screen 704 may be comprised external to but functionally connected with the apparatus 100 As used here, ‘connected’ means either ‘mechanically connected’ or ‘electrically connected’ either directly or indirectly. Connection does not have to be galvanic. Where the control system is concerned, connected means operably coupled to the extent that messages are transmitted and received via the appropriate communication means. The term “control system” may be understood to cover a controller, control module, or control element and need not necessary be a multi-element or distributed system (although it may be in some examples). It will be appreciated that various changes and modifications can be made to the present disclosed examples without departing from the scope of the present application as defined by the appended claims. Whilst endeavouring in the foregoing specification to draw attention to those features believed to be of particular importance it should be understood that the Applicant claims protection in respect of any patentable feature or combination of features hereinbefore referred to and / or shown in the drawings whether or not particular emphasis has been placed thereon.

Claims

1. An apparatus comprising:a processor,a memory configured to store computer-readable instructions that, when executed by the processor, cause the apparatus to:receive vocal data from a user, the vocal data indicative of a vocal response to a vocal exercise presented to the user, wherein the vocal exercise is configured to prompt the user to provide the vocal response indicative of a vocal characteristic;analyse the vocal data to determine a performance level of the vocal exercise, the performance level indicative of the vocal characteristic indicated in the vocal response compared to a target vocal characteristic;generate a further vocal exercise in dependence on the determined performance level and profile information of the user, wherein the profile information comprises an indication of content of interest to the user; andoutput the further vocal exercise.

2. The apparatus of claim 1, wherein the profile information of the user further comprises one or more of:an indication of a user objective corresponding to the vocal exercises; andan indication of a difficulty level.

3. The apparatus of any preceding claim, wherein the vocal exercise is determined based on the profile information of the user.

4. The apparatus of any preceding claim, wherein the apparatus is configured to generate the further vocal exercise to prompt the user to provide a further vocal response indicative of:the same vocal characteristic, ora different vocal characteristic,in dependence on the determined performance level.

5. The apparatus of any preceding claim, configured to:receive one or more previous vocal data from the user in response to a previous vocal exercise;analyse the one or more previous vocal data to determine a previous performance level of the previous vocal exercise; anddetermine the target vocal characteristic based on the previous performance level.

6. The apparatus of any preceding claim, wherein the vocal characteristic is one or more of: a maximum phonation, pitch range, jitter and shimmer, word intelligibility, sentence intelligibility, speech rate, a number of pauses, and an acoustic voice quality index.

7. The apparatus of any preceding claim, configured to compare the performance level to an upper threshold and a lower threshold, wherein:in the case that the performance level is above the upper threshold, increase a difficulty level of the vocal exercise with respect to the vocal characteristic associated with the performance level;in the case that the performance level is below the lower threshold, decrease the difficulty level of the vocal exercise with respect to the vocal characteristic associated with the performance level; andin the case that the performance level is below the upper threshold and above the lower threshold, maintain the difficulty level of the vocal exercise with respect to the vocal characteristic associated with the performance level.

8. The apparatus of any preceding claim, wherein the apparatus is configured to:receive a user input selecting a vocal exercise from a plurality of presented vocal exercises, wherein each vocal exercise has a different format for prompting the vocal response;determine a selection frequency for each vocal exercise in dependence on the user input; andstore the selection frequency in the profile information of the user, wherein the further vocal exercise is generated in dependence on the selection frequency.

9. The apparatus of any preceding claim, wherein, to generate the further vocal exercise, the apparatus is configured to:generate, using machine learning, one or more words based on the determined performance level and the profile information of the user;analyse the generated one or more words to determine a composition of the generated one or more words; andif the composition of the generated one or more words does not meet a predetermined criteria, re-generate the one or more words.

10. The apparatus of any preceding claim, configured to:generate a results user interface, the results user interface comprising an indication of the performance level to the user, wherein the results user interface comprises the target vocal characteristic mapped to the vocal data to indicate the performance level.

11. The apparatus of claim 10, wherein the target vocal characteristic is mapped to the vocal data by a colour coding, the colour of the colour coding being determined by the value of the performance level.

12. The apparatus of any preceding claim, wherein the apparatus is configured to determine the profile information in dependence on one or more of:a received user input comprising the indication of content of interest; anda learned user history.

13. A method for an apparatus, the method comprising:receiving vocal data from a user, the vocal data indicative of a vocal response to a vocal exercise presented to the user, wherein the vocal exercise is configured to prompt the user to provide the vocal response indicative of a vocal characteristic;analysing the vocal data to determine a performance level of the vocal exercise, the performance level indicative of the vocal characteristic indicated in the vocal response compared to a target vocal characteristic;generating a further vocal exercise in dependence on the determined performance level and profile information of the user, wherein the profile information comprises an indication of content of interest to the user; andoutputting the further vocal exercise.

14. The method of claim 13, comprising determining the vocal exercise based on the profile information of the user.

15. The method of any of claims 13 to 14, comprising generating the further vocal exercise to prompt the user to provide a further vocal response indicative of:the same vocal characteristic, ora different vocal characteristic,in dependence on the determined performance level.

16. The method of any of claims 13 to 15, comprising:receiving one or more previous vocal data from the user in response to a previous vocal exercise;analysing the one or more previous vocal data to determine a previous performance level of the previous vocal exercise; anddetermining the target vocal characteristic based on the previous performance level.

17. The method of any of claims 13 to 16 comprising comparing each of the performance level to an upper threshold and a lower threshold, wherein:in the case that the performance level is above the upper threshold, increasing a difficulty level with respect to the vocal characteristic associated with the performance level;in the case that the performance level is below the lower threshold, decreasing the difficulty level with respect to the vocal characteristic associated with the performance level; andin the case that the performance level is below the upper threshold and above the lower threshold, maintaining the difficulty level of the vocal exercise with respect to the vocal characteristic associated with the performance level.

18. The method of any of claims 13 to 17, comprising:receiving a user input selecting a vocal exercise from a plurality of vocal exercises, wherein each vocal exercise has a different format for prompting the vocal response;determining a selection frequency for each vocal exercise in dependence on the user input; andstoring the selection frequency in the profile information of the user, wherein the further vocal exercise is generated in dependence on the selection frequency.

19. The method of any of claims 13 to 18, when generating the further vocal exercise comprises:generating, using machine learning, one or more words based on the profile information of the user;analysing the generated one or more words to determine a composition of the generated one or more words; andif the composition of the generated one or more words does not meet a predetermined criteria, re-generating the one or more words.

20. The method of any of claims 13 to 19, comprising:generating a results user interface, the results user interface comprising an indication of the performance level to the user, wherein the results user interface comprises the target vocal characteristic mapped to the vocal data to indicate the performance level.

21. The method of claim 20, comprising:mapping the target vocal characteristic to the vocal data by a colour coding, the colour of the colour coding being determined by the value of the performance level.5 22. The method of any of claims 13 to 21, comprising determining the profile information independence on one or more of:a received user input comprising the indication of content of interest; anda learned user history.10 23. Computer program code which, when processed by a computer, configures thecomputer to perform the method of any of claims 13 to 22.

24. A system comprising:the apparatus of any of claims 1 to 12;15 a microphone configured to receive the vocal response; anda display screen configured to output the further vocal exercise.31

Images