Pascale Fung

Pascale Fung (馮雁) (born in Shanghai, China) is a co-founder and Chief Research and Innovation Officer of AMI Labs, an artificial intelligence research company focused on world models. She is a professor in the Department of Electronic & Computer Engineering and the Department of Computer Science & Engineering at the Hong Kong University of Science & Technology(HKUST). She is the director of the Centre for AI Research (CAiRE) at HKUST. She is an elected Fellow of the Institute of Electrical and Electronics Engineers (IEEE) for her “contributions to human-machine interactions”, an elected Fellow of the International Speech Communication Association for “fundamental contributions to the interdisciplinary area of spoken language human-machine interactions” and an elected Fellow of the Association for Computational Linguistics (ACL) for her “significant contributions toward statistical NLP, comparable corpora, and building intelligent systems that can understand and empathize with humans”. She is a member of the Global Future Council on Artificial Intelligence and Robotics, a think tank of the World Economic Forum, and blogs for the Forum's online publication Agenda. She is a member of the Partnership on AI. She has been invited as an AI expert to different government initiatives in China, Japan, the UAE, India, the European Union and the United Nations. Fung's publication topics include spoken language systems, natural language processing, and empathetic human-robot interaction. She co-founded the Human Language Technology Center (HLTC) and is an affiliated faculty with the Robotics Institute and the Big Data Institute, both at HKUST. Additionally, she is the founding chair of the Women Faculty Association at HKUST. She is actively involved in encouraging young women into careers in engineering and science. == Career and research interests == Fung's work is focused on building systems that try to understand and empathize with humans. She has authored and co-authored hundreds of publications, along with many journal listings and book chapters. Fung is often found in the media, among others as a writer for Scientific American, the World Economic Forum, and the London School of Economics, and the Design Society. She was a pioneer in using statistical models for natural language understanding. Her PhD thesis proposed unsupervised methods for aligning texts and mining dictionary translations in different languages by distributional properties. She is an expert in spoken language understanding and computer emotional intelligence, and is a strong proponent of technology transfer. Fung has applied many of her research group's results in the fields of, among others, robotics, IoT, and financial analytics. Her efforts led to the launch of the world's first Chinese natural language search engine in 2001, the first Chinese virtual assistant for smartphones in 2010, and the first emotional intelligent speaker in 2017. == Honors == Elected Fellow, Association for the Advancement of Artificial Intelligence (AAAI), for “significant contributions to the field of Conversational AI and to the development of ethical AI principles and algorithms” Elected Fellow, Association for Computational Linguistics (ACL), for “significant contributions toward statistical NLP, comparable corpora, and building intelligent systems that can understand and empathize with humans” Nominee, the VentureBeat AI Innovation Awards at Transform 2020, for "AI for Good" Awardee, 2017 Outstanding Women Professionals & Entrepreneurs Award, Hong Kong Women Professionals & Entrepreneurs Association Elected Fellow, Institute of Electrical and Electronics Engineers (IEEE), for “contributions to human-machine interactions” Elected Fellow, International Speech Communication Association (ISCA), for “fundamental contributions to the interdisciplinary area of spoken language human-machine interactions" Member, Global Future Council on AI and Robotics, World Economic Forum (2016–) One of the Top 50 Women of Hope, selected by List Magazine in 2014 Selected as “My Favorite Teacher” by top engineering students in 2007 and in 2009 == Affiliations == Fung is affiliated with the following institutions and organizations: Hong Kong University of Science and Technology World Economic Forum Institute of Electrical and Electronics Engineers Association for Computational Linguistics International Speech Communication Association Association for Computing Machinery Association for the Advancement of Artificial Intelligence

TikTok

TikTok is a social media and short-form online video platform. It hosts user-submitted videos, which range in duration from three seconds to 60 minutes. It can be accessed through a mobile app or through its website. Since its launch, TikTok has become one of the world's most popular social media platforms, using recommendation algorithms to connect content creators and influencers with new audiences. In April 2020, TikTok surpassed two billion mobile downloads worldwide. The popularity of TikTok has allowed viral trends in food, fashion, and music to take off and increase the platform's cultural impact worldwide. TikTok has come under scrutiny due to data privacy violations, mental health concerns, misinformation, offensive content, addictive algorithm, its role during the Gaza war, and, following its 2026 divestiture in the U.S., alleged censorship of criticism of Donald Trump and discussions of Jeffrey Epstein. While TikTok remains accessible to users in most countries, a minority of countries (including India and Afghanistan) have implemented full or partial bans. Many other countries limit TikTok's use on government-issued devices for security or privacy reasons. == Corporate structure == TikTok Ltd was incorporated in the Cayman Islands in the Caribbean and is based in both Singapore and Los Angeles. It owns entities which are based respectively in Australia (which also runs the New Zealand business), United Kingdom (also owns subsidiaries in the European Union), and Singapore (owns operations in Southeast Asia and India). A spin-off company, TikTok USDS Joint Venture LLC was formed on 22 January 2026 to handle TikTok and other ByteDance properties in the United States, Oracle Corporation, MGX Fund Management Limited, Silver Lake each holding a 15% stake, ByteDance holds a 19.9% stake and the remaining 35.1% is shared between Dell Technologies founder Michael Dell and Vastmere Strategic Investments. Its parent company, Beijing-based ByteDance, is owned by founders and Chinese investors, other global investors, and employees. One of ByteDance's main domestic subsidiaries is owned by Chinese state funds and entities through a 1% golden share. Employees have reported that multiple overlaps exist between TikTok and ByteDance in terms of personnel management and product development. TikTok says that since 2020, its US-based CEO is responsible for making important decisions, and has downplayed its China connection. == History == === Douyin === Douyin (Chinese: 抖音; pinyin: Dǒuyīn; lit. 'Shaking Sound') was launched on 20 September 2016, by ByteDance, originally under the name A.me, before changing its name to Douyin in December 2016. Douyin was developed in nearly 7 months and within a year had 100 million users, with more than one billion videos viewed every day. While TikTok and Douyin share a similar user interface, the platforms operate separately. Douyin includes an in-video search feature that can search by people's faces for more videos of them, along with other features such as buying, booking hotels, and making geo-tagged reviews. === TikTok === ByteDance planned on Douyin expanding overseas. The founder of ByteDance, Zhang Yiming, stated that "China is home to only one-fifth of Internet users globally. If we don't expand on a global scale, we are bound to lose to peers eyeing the four-fifths. So, going global is a must." ByteDance created TikTok as an overseas version of Douyin. TikTok was launched in the international market in September 2017. On 9 November 2017, ByteDance spent nearly $1 billion to purchase Musical.ly, a startup headquartered in Shanghai with an overseas office in Santa Monica, California. Musical.ly was a social media video platform that allowed users to create short lip-sync and comedy videos, initially released in August 2014. TikTok merged with Musical.ly on 2 August 2018 with existing accounts and data consolidated into one app, keeping the title TikTok. On 23 January 2018, the TikTok app ranked first among free application downloads on app stores in Thailand and other countries. TikTok has been downloaded more than 130 million times in the United States and has reached 2 billion downloads worldwide, according to data from mobile research firm Sensor Tower (those numbers exclude Android users in China). In the United States, Jimmy Fallon, Tony Hawk, and other celebrities began using the app in 2018. Other celebrities like Jennifer Lopez, Jessica Alba, Will Smith, and Justin Bieber joined TikTok. In January 2019, TikTok allowed creators to embed merchandise sale links into their videos. On 3 September 2019, TikTok and the US National Football League (NFL) announced a multi-year partnership. The agreement came just two days before the NFL's 100th season kick-off at Soldier Field in Chicago where TikTok hosted activities for fans in honor of the deal. The partnership entails the launch of an official NFL TikTok account, which is to bring about new marketing opportunities such as sponsored videos and hashtag challenges. In July 2020, TikTok, excluding Douyin, reported close to 800 million monthly active users worldwide after less than four years of existence. In May 2021, TikTok appointed Shou Zi Chew as their new CEO who assumed the position from interim CEO Vanessa Pappas, following the resignation of Kevin A. Mayer on 27 August 2020. In September 2021, TikTok reported that it had reached 1 billion users. In 2021, TikTok earned $4 billion in advertising revenue. In October 2022, TikTok was reported to be planning an expansion into the e-commerce market in the US, following the launch of TikTok Shop in the United Kingdom. The company posted job listings for staff for a series of order fulfillment centers in the US and was reportedly planning to start the new live shopping business before the end of the year. The Financial Times reported that TikTok will launch a video gaming channel, but the report was denied in a statement to Digiday, with TikTok instead aiming to be a social hub for the gaming community. According to data from app analytics group Sensor Tower, advertising on TikTok in the US grew by 11% in March 2023, with companies including Pepsi, DoorDash, Amazon, and Apple among the top spenders. According to estimates from research group Insider Intelligence, TikTok is projected to generate $14.15 billion in revenue in 2023, up from $9.89 billion in 2022. In March 2024, The Wall Street Journal reported that TikTok's growth in the US had stagnated. ==== Plans to sell TikTok's US operations ==== Since at least 2020, following calls to ban TikTok in the country, the Committee on Foreign Investment in the United States (CFIUS) has been investigating the company's 2017 merger with Musical.ly but has not finalized any of its negotiations with TikTok, such as the Project Texas proposal, waiting instead for Congress to act. In January 2025, Chinese officials began preliminary talks about potentially selling TikTok's US operations to Elon Musk if the app faced an impending ban due to national security concerns. While Beijing preferred TikTok remain under ByteDance's control, the sale could happen through a competitive process or with US government involvement. One possibility involved Musk's platform, X, taking over TikTok's US business. The move came ahead of a Supreme Court case that upheld the constitutionality of a law that would force a sale or ban of TikTok in the US by 19 January 2025, due to national security concerns regarding its ties to China. Other potential buyers included Project Liberty's "The People's Bid For TikTok" consortium of Frank McCourt with Kevin O'Leary, Steven Mnuchin, MrBeast and Bobby Kotick, the seriousness of these potential buyers was unclear. The day before the impending ban, California-based conversational search engine company Perplexity AI submitted a bid for a merger with TikTok US. On 14 September 2025, the Wall Street Journal reported the US and China have reached the "framework of a deal" for the US operations of TikTok to be sold to a consortium of investors in the US including close Trump ally Larry Ellison of Oracle. The deal was completed by 22 January 2026, with a consortium of investors—including Oracle, Silver Lake, MGX, and others including the personal investment entity for Michael Dell—owning more than 80% of the new venture. ByteDance retained 19.9% ownership. Under the deal, the app would remain the same, and the algorithm would be adjusted over time to favor American topics for those users. === Expansion in other markets === TikTok was downloaded over 104 million times on Apple's App Store during the first half of 2018, according to data provided to CNBC by Sensor Tower. After merging with musical.ly in August, downloads increased and TikTok subsequently became the most downloaded app in the US in October 2018, which musical.ly had done once before. In February 2019, TikTok, together with Douyin, hit one billion downloads globally, excluding Android

VACUUM

VACUUM is a set of normative guidance principles for achieving training and test dataset quality for structured datasets in data science and machine learning. The garbage-in, garbage out principle motivates a solution to the problem of data quality but does not offer a specific solution. Unlike the majority of the ad-hoc data quality assessment metrics often used by practitioners VACUUM specifies qualitative principles for data quality management and serves as a basis for defining more detailed quantitative metrics of data quality. VACUUM is an acronym that stands for: valid accurate consistent uniform unified model

Transduction (machine learning)

In logic, statistical inference, and supervised learning, transduction or transductive inference is reasoning from observed, specific (training) cases to specific (test) cases. In contrast, induction is reasoning from observed training cases to general rules, which are then applied to the test cases. The distinction is most interesting in cases where the predictions of the transductive model are not achievable by any inductive model. Note that this is caused by transductive inference on different test sets producing mutually inconsistent predictions. Transduction was introduced in a computer science context by Vladimir Vapnik in the 1990s, motivated by his view that transduction is preferable to induction since, according to him, induction requires solving a more general problem (inferring a function) before solving a more specific problem (computing outputs for new cases): "When solving a problem of interest, do not solve a more general problem as an intermediate step. Try to get the answer that you really need but not a more general one.". An example of learning which is not inductive would be in the case of binary classification, where the inputs tend to cluster in two groups. A large set of test inputs may help in finding the clusters, thus providing useful information about the classification labels. The same predictions would not be obtainable from a model which induces a function based only on the training cases. Some people may call this an example of the closely related semi-supervised learning, since Vapnik's motivation is quite different. The most well-known example of a case-bases learning algorithm is the k-nearest neighbor algorithm, which is related to transductive learning algorithms. Another example of an algorithm in this category is the Transductive Support Vector Machine (TSVM). A third possible motivation of transduction arises through the need to approximate. If exact inference is computationally prohibitive, one may at least try to make sure that the approximations are good at the test inputs. In this case, the test inputs could come from an arbitrary distribution (not necessarily related to the distribution of the training inputs), which wouldn't be allowed in semi-supervised learning. An example of an algorithm falling in this category is the Bayesian Committee Machine (BCM). == Historical context == The mode of inference from particulars to particulars, which Vapnik came to call transduction, was already distinguished from the mode of inference from particulars to generalizations in part III of the Cambridge philosopher and logician W.E. Johnson's 1924 textbook, Logic. In Johnson's work, the former mode was called 'eduction' and the latter was called 'induction'. Bruno de Finetti developed a purely subjective form of Bayesianism in which claims about objective chances could be translated into empirically respectable claims about subjective credences with respect to observables through exchangeability properties. An early statement of this view can be found in his 1937 La Prévision: ses Lois Logiques, ses Sources Subjectives and a mature statement in his 1970 Theory of Probability. Within de Finetti's subjective Bayesian framework, all inductive inference is ultimately inference from particulars to particulars. == Example problem == The following example problem contrasts some of the unique properties of transduction against induction. A collection of points is given, such that some of the points are labeled (A, B, or C), but most of the points are unlabeled (?). The goal is to predict appropriate labels for all of the unlabeled points. The inductive approach to solving this problem is to use the labeled points to train a supervised learning algorithm, and then have it predict labels for all of the unlabeled points. With this problem, however, the supervised learning algorithm will only have five labeled points to use as a basis for building a predictive model. It will certainly struggle to build a model that captures the structure of this data. For example, if a nearest-neighbor algorithm is used, then the points near the middle will be labeled "A" or "C", even though it is apparent that they belong to the same cluster as the point labeled "B", compared to semi-supervised learning. Transduction has the advantage of being able to consider all of the points, not just the labeled points, while performing the labeling task. In this case, transductive algorithms would label the unlabeled points according to the clusters to which they naturally belong. The points in the middle, therefore, would most likely be labeled "B", because they are packed very close to that cluster. An advantage of transduction is that it may be able to make better predictions with fewer labeled points, because it uses the natural breaks found in the unlabeled points. One disadvantage of transduction is that it builds no predictive model. If a previously unknown point is added to the set, the entire transductive algorithm would need to be repeated with all of the points in order to predict a label. This can be computationally expensive if the data is made available incrementally in a stream. Further, this might cause the predictions of some of the old points to change (which may be good or bad, depending on the application). A supervised learning algorithm, on the other hand, can label new points instantly, with very little computational cost. == Transduction algorithms == Transduction algorithms can be broadly divided into two categories: those that seek to assign discrete labels to unlabeled points, and those that seek to regress continuous labels for unlabeled points. Algorithms that seek to predict discrete labels tend to be derived by adding partial supervision to a clustering algorithm. Two classes of algorithms can be used: flat clustering and hierarchical clustering. The latter can be further subdivided into two categories: those that cluster by partitioning, and those that cluster by agglomerating. Algorithms that seek to predict continuous labels tend to be derived by adding partial supervision to a manifold learning algorithm. === Partitioning transduction === Partitioning transduction can be thought of as top-down transduction. It is a semi-supervised extension of partition-based clustering. It is typically performed as follows: Consider the set of all points to be one large partition. While any partition P contains two points with conflicting labels: Partition P into smaller partitions. For each partition P: Assign the same label to all of the points in P. Of course, any reasonable partitioning technique could be used with this algorithm. Max flow min cut partitioning schemes are very popular for this purpose. === Agglomerative transduction === Agglomerative transduction can be thought of as bottom-up transduction. It is a semi-supervised extension of agglomerative clustering. It is typically performed as follows: Compute the pair-wise distances, D, between all the points. Sort D in ascending order. Consider each point to be a cluster of size 1. For each pair of points {a,b} in D: If (a is unlabeled) or (b is unlabeled) or (a and b have the same label) Merge the two clusters that contain a and b. Label all points in the merged cluster with the same label. === Continuous Label Transduction === These methods seek to regress continuous labels, often via manifold learning techniques. The idea is to learn a low-dimensional representation of the data and infer values smoothly across the manifold. == Applications and related concepts == Transduction is closely related to: Semi-supervised learning – uses both labeled and unlabeled data but typically induces a model. Case-based reasoning – such as the k-nearest neighbor (k-NN) algorithm, often considered a transductive method. Transductive Support Vector Machines (TSVM) – extend standard SVMs to incorporate unlabeled test data during training. Bayesian Committee Machine (BCM) – an approximation method that makes transductive predictions when exact inference is too costly.

Ericom Connect

Ericom Connect is a remote access/application publishing solution produced by Ericom Software that provides secure, centrally managed access to physical or hosted desktops and applications running on Microsoft Windows and Linux systems. == Product overview == Ericom Connect is desktop virtualization and application virtualization software that allows users to run applications remotely, without installing them on the local computer or device. The software is noted for its scalability, ease of deployment, and compatibility with any type of infrastructure, cloud or physical. Ericom Connect uses AccessPad (native client for desktops), AccessToGo (native client for mobile), or AccessNow, one of the first HTML5 RDP solutions to support clientless access to Windows desktops and applications from any device with an HTML5-compatible browser, including Macintosh computers, mobile devices, and Google Chromebooks. Other notable features include performance monitoring, built-in real-time analytics & BI, support for two-factor authentication (using RSA SecurID), multi-tenancy and multi-datacenter support via a single unified web interface, and a “Launch Simulation” feature that allows users to visualize and simulate actual step-by-step user processes directly from within the administration console. In addition to scalability, by distributing configurations, logs, etc., across multiple servers there is no single point of failure, as can be the case if all configuration information is stored on one server. == History == Ericom Connect was introduced in 2015. Ericom Connect is a successor to Ericom PowerTerm Web Connect. PowerTerm Web Connect used an architecture similar to what was then current with Citrix and VMWare, relying on a centralized SQL server, a connection broker, image management for different hypervisors, and a variety of clients. Ericom Connect uses a new grid architecture that provides more scalability, reliability, and flexibility than before.

Network Abstraction Layer

The Network Abstraction Layer (NAL) is a part of the H.264/AVC and HEVC video coding standards. The main goal of the NAL is the provision of a "network-friendly" video representation addressing "conversational" (video telephony) and "non conversational" (storage, broadcast, or streaming) applications. NAL has achieved a significant improvement in application flexibility relative to prior video coding standards. == Introduction == An increasing number of services and growing popularity of high definition TV are creating greater needs for higher coding efficiency. Moreover, other transmission media such as cable modem, xDSL, or UMTS offer much lower data rates than broadcast channels, and enhanced coding efficiency can enable the transmission of more video channels or higher quality video representations within existing digital transmission capacities. Video coding for telecommunication applications has diversified from ISDN and T1/E1 service to embrace PSTN, mobile wireless networks, and LAN/Internet network delivery. Throughout this evolution, continued efforts have been made to maximize coding efficiency while dealing with the diversification of network types and their characteristic formatting and loss/error robustness requirements. The H.264/AVC and HEVC standards are designed for technical solutions including areas like broadcasting (over cable, satellite, cable modem, DSL, terrestrial, etc.) interactive or serial storage on optical and magnetic devices, conversational services, video-on-demand or multimedia streaming, multimedia messaging services, etc. Moreover, new applications may be deployed over existing and future networks. This raises the question about how to handle this variety of applications and networks. To address this need for flexibility and customizability, the design covers a NAL that formats the Video Coding Layer (VCL) representation of the video and provides header information in a manner appropriate for conveyance by a variety of transport layers or storage media. The NAL is designed in order to provide "network friendliness" to enable simple and effective customization of the use of VCL for a broad variety of systems. The NAL facilitates the ability to map VCL data to transport layers such as: RTP/IP for any kind of real-time wire-line and wireless Internet services. File formats, e.g., ISO MP4 for storage and MMS. H.32X for wireline and wireless conversational services. MPEG-2 systems for broadcasting services, etc. The full degree of customization of the video content to fit the needs of each particular application is outside the scope of the video coding standardization effort, but the design of the NAL anticipates a variety of such mappings. Some key concepts of the NAL are NAL units, byte stream, and packet formats uses of NAL units, parameter sets, and access units. A short description of these concepts is given below. == NAL units == The coded video data is organized into NAL units, each of which is effectively a packet that contains an integer number of bytes. The first byte of each H.264/AVC NAL unit is a header byte that contains an indication of the type of data in the NAL unit. For HEVC the header was extended to two bytes. All the remaining bytes contain payload data of the type indicated by the header. The NAL unit structure definition specifies a generic format for use in both packet-oriented and bitstream-oriented transport systems, and a series of NAL units generated by an encoder is referred to as a NAL unit stream. == NAL Units in Byte-Stream Format Use == Some systems require delivery of the entire or partial NAL unit stream as an ordered stream of bytes or bits within which the locations of NAL unit boundaries need to be identifiable from patterns within the coded data itself. For use in such systems, the H.264/AVC and HEVC specifications define a byte stream format. In the byte stream format, each NAL unit is prefixed by a specific pattern of three bytes called a start code prefix. The boundaries of the NAL unit can then be identified by searching the coded data for the unique start code prefix pattern. The use of emulation prevention bytes guarantees that start code prefixes are unique identifiers of the start of a new NAL unit. A small amount of additional data (one byte per video picture) is also added to allow decoders that operate in systems that provide streams of bits without alignment to byte boundaries to recover the necessary alignment from the data in the stream. Additional data can also be inserted in the byte stream format that allows expansion of the amount of data to be sent and can aid in achieving more rapid byte alignment recovery, if desired. == NAL Units in Packet-Transport System Use == In other systems (e.g., IP/RTP systems), the coded data is carried in packets that are framed by the system transport protocol, and identification of the boundaries of NAL units within the packets can be established without use of start code prefix patterns. In such systems, the inclusion of start code prefixes in the data would be a waste of data carrying capacity, so instead the NAL units can be carried in data packets without start code prefixes. == VCL and Non-VCL NAL Units == NAL units are classified into VCL and non-VCL NAL units. VCL NAL units contain the data that represents the values of the samples in the video pictures. Non-VCL NAL units contain any associated additional information such as parameter sets (important header data that can apply to a large number of VCL NAL units) and supplemental enhancement information (timing information and other supplemental data that may enhance usability of the decoded video signal but are not necessary for decoding the values of the samples in the video pictures). == Parameter Sets == A parameter set contains shared configuration data that is carried in non-VCL NAL units. Parameter sets are typically reused when decoding many coded pictures within a video sequence. Each VCL NAL unit references a picture parameter set (PPS), which in turn references a sequence parameter set (SPS). There are two types of parameter sets: Sequence parameter set (SPS), which specifies mostly constant configuration such as resolution, bit depth, or chroma format. (For a concrete implementation, see FFmpeg's SPS struct.) Picture parameter set (PPS), which applies on top of an SPS, and specifies configuration such as QP offsets. (For a concrete implementation, see FFmpeg's PPS struct.) The sequence and picture parameter-set mechanism decouples the transmission of infrequently changing information from the transmission of coded representations of the values of the samples in the video pictures. Each VCL NAL unit contains an identifier that refers to the content of the relevant picture parameter set and each picture parameter set contains an identifier that refers to the content of the relevant sequence parameter set. In this manner, a small amount of data (the identifier) can be used to refer to a larger amount of information (the parameter set) without repeating that information within each VCL NAL unit. Sequence and picture parameter sets can be sent well ahead of the VCL NAL units that they apply to, and can be repeated to provide robustness against data loss. In some applications, parameter sets may be sent within the channel that carries the VCL NAL units (termed "in-band" transmission). In other applications, it can be advantageous to convey the parameter sets "out-of-band" using a more reliable transport mechanism than the video channel itself. == Access Units == A set of NAL units in a specified form is referred to as an access unit. The decoding of each access unit results in one decoded picture. Each access unit contains a set of VCL NAL units that together compose a primary coded picture. It may also be prefixed with an access unit delimiter to aid in locating the start of the access unit. Some supplemental enhancement information containing data such as picture timing information may also precede the primary coded picture. The primary coded picture consists of a set of VCL NAL units consisting of slices or slice data partitions that represent the samples of the video picture. Following the primary coded picture may be some additional VCL NAL units that contain redundant representations of areas of the same video picture. These are referred to as redundant coded pictures, and are available for use by a decoder in recovering from loss or corruption of the data in the primary coded pictures. Decoders are not required to decode redundant coded pictures if they are present. Finally, if the coded picture is the last picture of a coded video sequence (a sequence of pictures that is independently decodable and uses only one sequence parameter set), an end of sequence NAL unit may be present to indicate the end of the sequence; and if the coded picture is the last coded picture in the entire NAL unit stream, an end of stream NAL unit may be present to

Pronunciation assessment

Automatic pronunciation assessment uses computer speech recognition to determine how accurately speech has been pronounced, instead of relying on a human instructor or proctor. It is also called speech verification, pronunciation evaluation, and pronunciation scoring. This technology is used to grade speech quality, for language testing, for computer-aided pronunciation teaching (CAPT) in computer-assisted language learning (CALL), for speaking skill remediation, and for accent reduction. Pronunciation assessment is different from dictation or automatic transcription, because instead of determining unknown speech, it verifies learners' pronunciation of known word(s), often from prior transcription of the same utterance; ideally scoring the intelligibility of the learners' speech. Sometimes pronunciation assessment evaluates the prosody of the learners' speech, such as intonation, pitch, tempo, rhythm, and syllable and word stress, although those are usually not essential for being understood in most languages. Pronunciation assessment is also used in reading tutoring, for example in products from Google, Microsoft, and Amira Learning. Automatic pronunciation assessment can also be used to help diagnose and treat speech disorders such as apraxia. == Intelligibility == Intelligibility refers to how well a learner's utterance is understood by a listener, rather than how much it sounds like a native speaker. This is separate from measures of fluency, such as so-called "Goodness of Pronunciation" (GoP) scores, which estimate how closely an utterance aligns with those of native speakers. Intelligibility is widely regarded as the most important communicative goal in pronunciation teaching and assessment. For example, in the Common European Framework of Reference for Languages (CEFR) assessment criteria for "overall phonological control", intelligibility outweighs formally correct pronunciation at all levels. Studies in applied linguistics have shown that accent reduction does not always increase intelligibility because listeners can often comprehend heavily accented speech without difficulty. Pronunciation assessment systems often rely on acoustic methods such as GoP which compare learner speech to reference models to produce phoneme-level scores, which are in turn aggregated to produce word and phrase scores. While these methods are effective for identifying deviations from native speakers' utterances, they do not effectively measure how understandable speech is to human listeners. Intelligibility is influenced by broader linguistic and contextual factors such as stress placement, speech rate, and coarticulation, which are not represented in purely segmental scores. The earliest work on pronunciation assessment avoided measuring genuine listener intelligibility, a shortcoming corrected in 2011 at the Toyohashi University of Technology, and included in the Versant high-stakes English fluency assessment from Pearson and mobile apps from 17zuoye Education & Technology, but still missing in 2023 products from Google Search, Microsoft, Educational Testing Service, Speechace, and ELSA. Assessing authentic listener intelligibility is essential for avoiding inaccuracies from accent bias, especially in high-stakes assessments; from words with multiple correct pronunciations; and from phoneme coding errors in machine-readable pronunciation dictionaries. In 2022, researchers found that some newer speech-to-text systems, based on end-to-end reinforcement learning to map audio signals directly into words, produce word and phrase confidence scores (from 10-25ms audio frame logit aggregation) closely correlated with genuine listener intelligibility. Others have been able to assess intelligibility using Levenshtein or dynamic time warping distance measures from Wav2Vec2 representation of good speech. Further work through 2025 has focused specifically on measuring intelligibility. A 2025 study of 42 pronunciation and speech coaching apps (32 mobile and 10 web) found that none offered intelligibility assessment. Instead, most provided only segmental and accent-focused scoring. About two-thirds of the apps provided some form of specific pronunciation feedback, usually with phonetic transcriptions, but accompanied by visual cues (such as animations of the vocal tract or the lips and tongue from the front) in only about 5% of the apps. Less than a third provided feedback on learner perception of exemplar speech. == Evaluation == Although there are as yet no industry-standard benchmarks for evaluating pronunciation assessment accuracy, researchers occasionally release evaluation speech corpuses for others to use for improving assessment quality. Such evaluation databases often emphasize formally unaccented pronunciation to the exclusion of genuine intelligibility evident from blinded listener transcriptions. As of mid-2025, state of the art approaches for automatically transcribing phonemes typically achieve an error rate of about 10% from known good speech. The International Speech Communication Association (ISCA) 2025 Workshop on Speech and Language Technology in Education (SLaTE) administered a Speak & Improve Challenge: Spoken Language Assessment and Feedback, introducing benchmarks for evaluating pronunciation assessment and remediation systems across languages, accents, and learner populations. The challenge emphasized cross-lingual generalization and alignment with human intelligibility judgments, for more robust and interpretable assessment systems. Ethical issues in pronunciation assessment are present in both human and automatic methods. Authentic validity, fairness, and mitigating bias in evaluation are all crucial. Diverse speech data should be included in automatic pronunciation assessment models. Combining human judgments, especially blinded transcriptions from a wide diversity of listeners, with automated feedback can improve accuracy and fairness. Second language learners benefit substantially from their use of widely available speech recognition systems for dictation, virtual assistants, and AI chatbots. In such systems, users naturally try to correct their own errors evident in speech recognition results that they notice. Such use improves their grammar and vocabulary development along with their pronunciation skills. The extent to which explicit pronunciation assessment and remediation approaches improve on such self-directed interactions remains an open question. Similarly, automatic dictation results have been shown to reflect intelligibility about as well as human scorers. == Recent developments == During 2021–22, a smartphone-based CAPT system was used to sense articulation through both audible and inaudible signals, providing feedback at the phoneme level. Some promising areas for improvement which were being developed in 2024 include articulatory feature extraction and transfer learning to suppress unnecessary corrections. Other interesting advances under development include "augmented reality" interfaces for mobile devices using optical character recognition to provide pronunciation training on text found in user environments. In 2024, audio multimodal large language models were first described as assessing pronunciation. That work has been carried forward by other researchers in 2025 who report positive results. Subsequently, researchers demonstrated pronunciation scoring by providing a language model with textual descriptions of speech, including the speech-to-text transcript, phoneme sequences, pauses, and phoneme sequence matching; this approach can achieve performance similar to multimodal LLMs that analyze raw audio while avoiding their higher computational cost. In 2025, the Duolingo English Test authors published a description of their pronunciation assessment method, purportedly built to measure intelligibility rather than accent imitation. While achieving a correlation of 0.82 with expert human ratings, very close to inter-rater agreement and outperforming alternative methods, the method is nonetheless based on experts' scores along the six-point CEFR common reference levels scale, instead of actual blinded listener transcriptions. Further promising work in 2025 includes assessment feedback aligning learner speech to synthetic utterances using interpretable features, identifying continuous spans of words for remediation feedback; synthesizing corrected speech matching learners' self-perceived voices, which they prefer and imitate more accurately as corrections; and streaming such interactions. On January 21, 2026, Educational Testing Service's TOEFL iBT high-stakes English language test, required by US university admissions and employers from English as a foreign language applicants more often than all other internet-based tests combined, changed its speaking assessments. While official rubrics claim that the new scoring will be based primarily on intelligibility, the new test's technical description indicates that it ju