hash stringlengths 32 32 | doc_id stringlengths 7 13 | section stringlengths 3 121 | content stringlengths 0 2.2M |
|---|---|---|---|
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.1.2.5.19 Standard Hardware and Software Features for the CPE | QoS capabilities: Advanced Firewall Features • Network Address Translation (NAT). • Stateful Packet Inspection (SPI). • VPN Pass-through/Multi-sessions PPTP/L2TP/IPSec. Standards: • IEEE 802.11n™ [35]. • IEEE 802.11g™ [36]. • IEEE 802.3™ [37]. • IEEE 802.3u™ [38]. Switching Protocol • Ethernet - fast switching. • Perfo... |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.1.2.6 Performance | |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.1.2.6.1 Forwarding | CPE tunnel forwarding performance has to not be lower than 75 % of total connected downstream bandwidth and 40 % of total connected upstream bandwidth. |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.1.2.6.2 Node latency | Node latency has to be under 2 ms. |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.1.2.6.3 Max sessions | The CPE router has to be able to hold 10 000 sessions minimum. |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.1.2.7 DS-Lite | |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.1.2.7.1 General | Dual-Stack Lite IETF RFC 6333 [17] enables both continued support for IPv4 services and incentives for the deployment of IPv6. It also de-couples IPv6 deployment in the service provider network from the rest of the Internet, making incremental deployment easier. Dual-Stack Lite enables a broadband service provider to s... |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.1.2.7.2 WAN requirements: | The CE router has to support configuration of DS-Lite via the DS-Lite DHCPv6 option IETF RFC 6334 [20]. The IPv6 CE router may use other mechanisms to configure DS-Lite parameters. Such other mechanisms are not specified within the present document. The IPv6 CE router has to be engineered so that it does not perform IP... |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.1.2.7.3 Security Considerations | It is best practice to filter malicious traffic (e.g. spoofed packets, "Martian" addresses, etc.). Thus, the IPv6 CE router should support basic stateless egress and ingress filters. The Cabe router is also expected to offer mechanisms to filter traffic entering the customer network; however, the method by which vendor... |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.1.2.7.4 Security requirements: | The IPv6 CE router should support IETF RFC 6092 [31]. In particular, the IPv6 CE router should support functionality sufficient for implementing the set of recommendations in IETF RFC 6092 [31], section 4. Such functionality could be enabled by default or mechanisms by which users would configure it. The IPv6 CE router... |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.2 DHCP | |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.2.1 DHCPv6 - PD Internal and external | DHCPv6 prefix delegation for the CPE WAN interface has to be supported. The CPE has to also be able to distribute an EUI-64 allocation with future extension to EUI-112. It would be further beneficial if it is engineered against the latest version of the DHCPv6 specifications, to allocate stateful node addresses without... |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.3 DNS over IPv6 | DNS has to be able to be transported over both IPv4 and IPv6. |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.4 Client Reactivity | Within the remit of traffic requiring DS-Lite function, decapsulation, encapsulation and NAT function, the AFTR has to perform its function from ingress interface to egress interface in a time measured no higher than 200 microseconds under 50 % load. This includes requirements for fragmentation, reassembly, re-ordering... |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.5 Monitoring and Logging | The DS-Lite AFTR, being an inline device and typically deployed either centralized or decentralized and thus forming part of the operator's network, will be operated and monitored like any other device in the network. For this purpose CLI access could be allowed through either Telnet or SSH, the latter of which decreas... |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.6 Resource Management | The AFTR has to have the ability to share or split resources where possible to allow for full functionality in any given topology. This would include a single address per chassis. NPUs have to be able to share a single pool of address space on IPv4 and DHCP pool handoff on IPv6 if the function is required. Memory, proc... |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.7 Placement of function on LSN or CPE | DS-Lite implementation is based upon solely business continuity, i.e. how to give current and new customers access to their current service with minimal service change or impact as possible. Two technologies may be adopted to support this requirements, DS-Lite and 6to4, with DS-Lite being the primary utilization for In... |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.8 Security | |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.8.1 SNMP | |
3c4eaa0df36fcc7906df2e1cc795abea | 103 443-3 | 10.8.1.1 SNMP Management | SNMP has to be able to be transported over both IPv4 as well as IPv6 as no assumptions can be made over the operator's connectivity and ability to handle either AF in their management network. For increased security SNMPv3 has to be supported however not exclusively; SNMP v2c and v1 has to also be supported. ETSI ETSI ... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 1 Scope | The present document defines test methods, performance metrics and requirements for the voice assistance functionality of devices, including devices meant to be used in far talk conditions. The methods include definition of: input speech signals, positional relations of talkers and devices; acoustic environment charact... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 2 References | |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 2.1 Normative references | References are either specific (identified by date of publication and/or edition number or version number) or non-specific. For specific references, only the cited version applies. For non-specific references, the latest version of the referenced document (including any amendments) applies. Referenced documents which a... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 2.2 Informative references | References are either specific (identified by date of publication and/or edition number or version number) or non-specific. For specific references, only the cited version applies. For non-specific references, the latest version of the referenced document (including any amendments) applies. NOTE: While any hyperlinks i... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 3 Definition of terms, symbols and abbreviations | |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 3.1 Terms | For the purposes of the present document, the following terms apply: close talk: voice-based interaction between the voice assistant device and the user at a distance lesser than 30 cm far talk: voice-based interaction between the voice assistant device and the user at a distance greater than 2 m near talk: voice-based... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 3.2 Symbols | For the purposes of the present document, the following symbols apply: dB decibel dBFS dB referenced to digital full scale dBPa dB referenced to 1 Pascal dBSPL dB referenced to sound pressure level (2e-5 Pascal) dBSPL(A) dBSPL A-weighted according to IEC 61672-1 [10] dBSPL(C) dBSPL C-weighted according to IEC 61672-1 [... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 3.3 Abbreviations | For the purposes of the present document, the following abbreviations apply: ASR Active Speech Recognition CUR Correct Utterance Ratio DRR Direct to Reverb Ratio DUT Device Under Test FAR False Acceptance Rate FRR False Rejection Rate HATS Head And Torso Simulator HVAC Heating Ventilation Air Conditioning LOS Line-Of-S... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4 Acoustic environments for testing | |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.1 Introduction | Voice assistant devices come in a variety of form factors and have many intended use cases. This clause defines several acoustic environments for testing, each of which may be more applicable to some devices than others. The indoor acoustic environment described in clause 4.2 (whether real, as described in clause 4.2.2... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.2 Indoor acoustic environments | |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.2.1 General considerations | |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.2.1.1 Introduction | The indoor acoustic environments described in this clause are applicable to devices, which are envisioned to be used inside the home (e.g. smart speakers, soundbars, etc.). Many voice assistant devices are designed for indoor use cases. Such devices are subject to a variety of reverberation characteristics and interfer... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.2.1.2 Reverberation characteristics | Reverberation is a source of variation in the performance of ASR systems [i.9]. The reverberation characteristics of indoor spaces can vary substantially depending on the treatment of surfaces and volume of the environment. ETSI ETSI TS 103 504 V1.1.1 (2020-07) 11 The most common parameter used to characterize the reve... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.2.1.3 Interfering background noise | Voice assistant devices are expected to function in the presence of day-to-day background noise. These interferences may impact the ability of a voice assistant device to detect wake words and/or accurately recognize a target request. Interfering background noises may be stationary or non-stationary with respect to bot... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.2.2 Home-like test environment | |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.2.2.1 Introduction | This clause defines criteria for a home-like test environment that can be used to emulate the acoustic environments in which voice assistant devices are commonly used. The goal of the home-like test environment is to provide a space with reasonable reproducibility and repeatability, while offering some level of flexibi... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.2.2.2 Floor plan and room requirements | This clause contains general requirements for the home-like test environment, which shall be simultaneously met. The test environment emulates an "open floor plan" apartment with a kitchen and a living room area. Figure 4.2.2.2-1 presents the home-like test environment floor plan. The placements and dimensions of furni... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.2.2.3 Reverberation characteristics | The home-like test environment shall allow for room adjustments to meet several acoustic conditions for testing. Figure 4.2.2.3-1 demonstrates an alternative layout where Rug 1 and Rug 2 are removed from the room and both sets of curtains are opened. This layout increases reverberation time compared to the layout prese... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.2.2.4 Interfering background noise | The home-like environment uses discrete noise sources representing typical noise sources that are found in a home environment. The discrete noise sources are positioned according to the layout in Figure 4.2.2.2-1. For the kitchen sink and interfering talker noise sources, near-field monitor loudspeakers with a 3" to 5"... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.2.3 Lab-based simulation of acoustic environments | |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.2.3.1 Introduction | This clause defines criteria for lab-based simulation of acoustic environments that can be used wherever high repeatability and reproducibility in testing are desired. Lab-based simulation shall feature both reverberation and background noise simulation systems for the realization of various reverberation characteristi... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.2.3.2 Room requirements | Test room requirements are described in ETSI TS 103 557 [2]. An example loudspeaker setup for a simulation arrangement for far talk devices is shown in Figure 4.2.3.2-1. Figure 4.2.3.2-1: Example loudspeaker setup for lab-based simulation of acoustic environments |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.2.3.3 Simulation of reverberation characteristics | The loudspeaker equalization/calibration process and reverberation simulation procedure in a lab-based test environment are defined in ETSI TS 103 557 [2]. ETSI ETSI TS 103 504 V1.1.1 (2020-07) 16 Clause A.2 of ETSI TS 103 557 [2] presents a database of multi-channel room impulse responses for the lab-based simulation ... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.2.3.4 Simulation of interfering background noise | The loudspeaker equalization/calibration process and background noise simulation procedure in a lab-type environment are described in ETSI TS 103 224 [1]. Clause 8.4 of ETSI TS 103 224 [1] presents a database of multi-channel recordings for the lab-based simulation of interfering background noise conditions presented i... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.3 Vehicular acoustic environments | The vehicular acoustic environments described in this clause are applicable to devices, which are envisioned to be used inside the car (e.g. voice-controlled navigation systems, smart dashboards, etc.). Multiple setups and procedures of the present document are also applicable to vehicle-mounted voice assistant devices... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.4 Generic acoustic environments | |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.4.1 Introduction | The indoor acoustic environment described in clause 4.2 (either real or simulated) represents the usage of voice assistant devices in an "open floor plan" home-like environment. The multiple talker and DUT positions and orientations provide a good insight into the device performance for everyday use inside the home. Ho... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.4.2 Room requirements | Test room requirements are identical to the ones described in clause 4.2.3.2. |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.4.3 Simulation of reverberation characteristics | The loudspeaker equalization/calibration process and reverberation simulation procedure in a lab-based test environment are defined in clause A.1 of ETSI TS 103 557 [2]. These conditions use the asymmetric microphone array configuration according to ETSI TS 103 224 [1]. DUTs shall be tested with simulated reverberation... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.4.4 Simulation of interfering background noise | The loudspeaker equalization/calibration process and background noise simulation procedure for the generic test environments are included in clause 8 (Table 8.2) of ETSI TS 103 224 [1]. The noise types according to Table 4.4.4-1 shall be used for testing. These recordings were made in several acoustic environments whic... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 4.4.5 Definition of generic conditions | Table 4.4.5-1 defines meaningful combinations of simulated noise conditions (see Table 4.4.4-1) and reverberation scenarios (see Table 4.4.3-1). For measurements without noise simulation, i.e. under silence conditions, no noise type is provided. For measurements without simulation of reverberation, no scenario is provi... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5 Test setup and configurations | |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.1 Desired talker | |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.1.1 Introduction | For testing of voice assistant devices, either actual subjects, a HATS, or a standalone mouth simulator shall be used as the desired talker sound source. Voice assistant testing involves extensive repeat of wake words and/or voice commands, for which the use of pre-recorded speech materials reproduced with a HATS/mouth... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.1.2 Mouth directivity and orientation | Reproduction directivity is an important consideration when choosing an appropriate loudspeaker to emulate humans as sound sources. Humans have various directivity characteristics that exercise the actual acoustic propagation characteristics that a device will be exposed to in practical use. For example, sound source d... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.1.3 HATS/mouth simulator setup | For both home-like and lab-based test environments, a HATS, according to Recommendation ITU-T P.58 [3], or a mouth simulator, according to Recommendation ITU-T P.51 [8], shall be used. Based on the improved directivity characteristics demonstrated in annex D, a HATS should be used. Home-like test environment • HATS/mou... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.1.4 Speech material and level | |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.1.4.1 Introduction | When testing the automatic speech recognition functionality of a voice assistant device, the speech test materials shall include different types of suitable utterances/words/sentences, depending on the DUT/type of measurement, representing typical commands that are expected to be uttered to the device. The speech mater... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.1.4.2 Speech sound pressure level | The wake word and question shall be individually normalized to a nominal active speech level of -1,7 dBPa at the MRP for quiet conditions (e.g. those conditions without interfering background noise or barge-in content). This level is equivalent to the typical speech level at the MRP for speakerphone hands-free terminal... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.1.4.3 Command pause | The command pause is a short gap in speech between the wake word and the command. The length of the command pause may be adjusted to assess the performance of the device under different pause conditions. The length of the pause shall be indicated in the test report. |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.1.4.4 Number of talkers | A minimum of 10 different talkers, representing expected device users, should be used for testing. The distribution of talkers shall be balanced between male and female talkers, spanning a wide range of talker's pitch. |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.1.4.5 Requirements on speech materials recordings | The recording of speech materials shall be conducted in a quiet and mostly anechoic environment, with T60 < 0,15 s from 200 Hz to 8 kHz. The recording room ambient noise, microphone self-noise and acquisition hardware self-noise shall be capable of meeting an NC 20 or lower noise rating measured according to ANSI/ASA S... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.1.4.6 Other considerations on speech materials | Other aspects for consideration and reporting (recommended, but not mandatory): • The different talkers should have either no or the same dialect, different dialects should not be mixed. The dialect used for one evaluation shall be reported. Any other pertinent linguistic or talker details should be provided. • In case... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.2 Voice assistant device setup | |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.2.1 Introduction | The setup of the voice assistant device with respect to the talker, in combination with the acoustic characteristics of the environment and noise level, determines the Direct to Reverb Ratio (DRR) and Signal to Noise Ratio (SNR) experienced by the voice assistant device. The DRR and SNR are important parameters to defi... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.2.2 Device positioning in the test environment | Home-like test environment Placement of the DUT is an important factor to be considered in testing. In a lab environment, it is often common to setup DUTs in the centre of the room for testing. However, most voice assistant devices are used at home in a location close to a wall outlet. Since DUTs feature multi-micropho... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.2.3 Device orientation | The voice assistant device shall be tested with different device orientations, unless the device is expected to be used solely at a given orientation (e.g. a soundbar intended to be wall mounted). For devices that are not intended to be used solely at a given orientation, three orientations shall be tested as per Table... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.2.4 Line-of-sight issues | The voice assistant device shall be tested both with and without direct Line-Of-Sight (LOS) to the desired talker. Line-of-sight issues are potentially problematic when beamforming algorithms attempt to identify a dominant direction of arrival for the speech. Note that line-of-sight is not to be confused with the direc... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.3 Interaction tests | |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.3.1 Barge-in | |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.3.1.1 Introduction | Voice controlled devices can play back self-generated sounds (e.g. music) by integrated loudspeakers. The task of recording and playing back at the same time is similar to a double talk situation in telecommunication: the device has to carry out an echo cancellation and/or suppression of the known output signal to perf... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.3.1.2 Barge-in test conditions | Home-like test environment DUT and HATS/mouth simulator position: For the home-like test environment, the DUT shall be positioned at DUT Position 1 and Talker Position 3 and should be positioned at the other location combinations. • Speech level: The speech level for the desired talker shall be set according to clause ... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.3.2 Conversation mode | The device is asked a question and gives an answer. The DUT should answer an ambiguous second question using information from the first question. EXAMPLES: User: "How old is Tom Hanks?" Answer: "He is 61 years old." User: "Which movies did he play in?" Answer of DUT1: "His last movie was 'The circle'" (expected... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.3.3 Dialogue mode | If a device is tested for the completion of tasks, it is desired to have it completed successfully, reliably, quickly and without unnecessary additional interactions. The accuracy of each task can be expressed per talker (and possibly per varying intonations/synonyms) and is scored according to the scheme in Table 5.3.... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.4 Wake word detection tests | |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.4.1 False Rejection Rate (FRR) | False Rejection Rate (FRR) should be evaluated with at least 60 wake word trials for statistically significant results on a per condition basis. See annex C for more information on sample size recommendations for statistically significance in Bernoulli trials, for FRR and other performance metrics described in clause 6... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 5.4.2 False Acceptance Rate (FAR) | While devices should detect and respond to occurrences of their wake word, they should not mistake other speech for their wake word, which is a false acceptance (sometimes also referred as 'false alarm' or 'false detect'). To measure the False Acceptance Rate (FAR), the DUT shall be exposed to 24 hours of generic audio... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 6 Performance metrics | |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 6.1 Introduction | The test procedures described in clause 5 can be carried out over a large number of variables. The metrics introduced in this clause are described per utterance/sentence, per talker and/or per condition. In this context, condition means a specific combination of the variables: • Distance/Playback Level. • Background no... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 6.2 Word Error Rate (WER) | To evaluate this metric, the written transcript shall be available for each source utterance. It can directly be used for the calculation of the word error rate WERi,j,k (corresponding to the i-th utterance of the j-th talker in the k-th condition) according to equation 6.2-1. , , = , , , , , , , , (6.2-1) Here... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 6.3 False Rejection Rate (FRR) | The FRR of the j-th talker and k-th condition is calculated according to equation 6.3-1 as the ratio between the number of wake words missed, , , and the total number of wake word trials, , . , = , , (6.3-1) NOTE: See annex C for more information on sample size recommendations for statistically significa... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 6.4 False Acceptance Rate (FAR) | The FAR is calculated according to equation 6.4-1 as the number of false alarms, ±, divided by the observation time. The FAR is reported in false alarms per unit time. = ± (6.4-1) |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 6.5 Task Completion Rate (TCR) | For the evaluation of dialogue systems according to clause 5.3.3, each task is evaluated for the i-th utterance (e.g. multiple modifications for the same task), the j-th talker and k-th condition according to Table 5.3.3-1, providing a single result Vi,j,k for each trial. The Task Completion Rate (TCR) per condition is... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 6.6 Task Completion Time (TCT) | The Task Completion Time (TCT) is presented in Figure 6.1-1 as the time elapsed between when the DUT completes recording the question (step 7), ,, , and when the DUT finishes processing the question and generating the response (step 8), ,, . The TCT per condition k is calculated for all utterances i and tal... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 6.7 Perceived Task Completion Time (PTCT) | The Perceived Task Completion Time (PTCT) is presented in Figure 6.1-1 as the time elapsed between when the talker completes the question utterance (step 5), ,, , and when the DUT begins its response (step 9), ,, . The PTCT per condition k is calculated for all utterances i and talkers j according to equ... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 6.8 Wake Word Delay (WWD) | The Wake Word Delay (WWD) is presented in Figure 6.1-1 as the time elapsed between when the DUT completes recording the wake word (step 2), ,, , and when the DUT flags that the wake word has been detected (step 3), ,, . WWD per condition k is calculated for utterance i and talker j according to equation ... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 6.9 Perceived Wake Word Delay (PWWD) | The Perceived Wake Word Delay (PWWD) is presented in Figure 6.1-1 as the time elapsed between when the talker completes uttering the wake word (step 1), ,, , and when the DUT indicates that the wake word has been detected (step 4), ,, . PWWD per condition k is calculated for utterance i and talker j acco... |
95bdd4f9e3c21c512c868e3b7710fe98 | 103 504 | 7 Performance requirements | The definition of performance requirements for the previously described metrics is open for further study. ETSI ETSI TS 103 504 V1.1.1 (2020-07) 33 Annex A (normative): Room acoustics and electro acoustic equipment positioning The positioning of transducers in the acoustic environment can strongly influence their effec... |
e16d334f11519b9671e26952047d356b | 103 491 | 1 Scope | The present document describes a compressed audio delivery system comprised of a metadata component and a compressed audio component. This system is referred to as DTS-UHD. The stream and metadata structures described here allow efficient delivery of immersive audio. Additionally, the bitstream format of a new audio co... |
e16d334f11519b9671e26952047d356b | 103 491 | 2 References | |
e16d334f11519b9671e26952047d356b | 103 491 | 2.1 Normative references | References are either specific (identified by date of publication and/or edition number or version number) or non-specific. For specific references, only the cited version applies. For non-specific references, the latest version of the referenced document (including any amendments) applies. Referenced documents which a... |
e16d334f11519b9671e26952047d356b | 103 491 | 2.2 Informative references | References are either specific (identified by date of publication and/or edition number or version number) or non-specific. For specific references, only the cited version applies. For non-specific references, the latest version of the referenced document (including any amendments) applies. NOTE: While any hyperlinks i... |
e16d334f11519b9671e26952047d356b | 103 491 | 3.1 Terms | For the purposes of the present document, the following terms apply: ACE channel: one waveform within an ACE Stream ACE frame: one ACE access unit ACE group of frames: contiguous sequence of frames that start with a Sync Frame ACE sync frame: ACE frame that may also serve as a random access point ACE stream: subset of ... |
e16d334f11519b9671e26952047d356b | 103 491 | 3.2 Symbols | Void. |
e16d334f11519b9671e26952047d356b | 103 491 | 3.3 Abbreviations | For the purposes of the present document, the following abbreviations apply: ACE Audio Compression Engine ACN Ambisonic Channel Number AM Adapter Matrix API Application Program Interface ATSC Advanced Television Systems Committee CBA Channel-Based Audio CLR Center Left Right CRC Cycle Redundancy Check DASH Dynamic Adap... |
e16d334f11519b9671e26952047d356b | 103 491 | 3.4 Document Conventions | A number of conventions are applied throughout the present document: • In parameter descriptions, most significant byte and most significant bit first (big endian), convention is utilized unless otherwise specified. • The existence of many bit fields is determined by conditions of prior bits in the stream. As such, in ... |
e16d334f11519b9671e26952047d356b | 103 491 | 4 DTS-UHD System Overview | |
e16d334f11519b9671e26952047d356b | 103 491 | 4.1 Overview | The DTS-UHD coding system is the third generation of DTS audio delivery formats. It is designed to both improve efficiency and deliver a richer set of features than the second generation DTS system. The first two generations of DTS codecs were designed primarily for Channel-Based Audio (CBA). A primary advantage of CBA... |
e16d334f11519b9671e26952047d356b | 103 491 | 4.2 Stream Construction |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.