Datasets:

CHiME9-ECHI
/

CHiME9-ECHI

Name: CHiME9-ECHI
Creator: ECHI

Languages:

English

DOI:

Dataset card Files Files and versions

xet

Community

You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

CHiME-9 ECHI Data Use Agreement

Dataset Version: June 2025
Maintained by: CHiME-9 ECHI Challenge Organizers
Contact: echichallenge@gmail.com

Overview

The CHiME-9 ECHI Dataset contains multi-speaker conversation recordings in cafeteria-like noise conditions collected over hearing aid and smart glasses devices under natural conversational settings. Due to the sensitive and identifiable nature of the data, its use is subject to the following terms and conditions.

By accessing or using the CHiME-9 ECHI Dataset, you acknowledge that you have read, understood, and agreed to comply with the terms below.

1. License Rights and Permitted Use

a. Grant of Rights

You are granted a non-exclusive, non-transferable, worldwide, royalty-free license to use the CHiME-9 ECHI Dataset solely for:

Academic and educational purposes
Algorithm development and benchmarking
Publication of models, visualizations, or results, provided proper attribution is given (see Section 3)

b. Prohibited Use

You may not, without prior written permission:

Redistribute or publicly repost any part of the raw dataset (video, audio, transcripts, face landmarks) with the exception of for purposes of showcasing your results achieved on the Dataset with proper attribution (see below).
Upload the dataset to public repositories, cloud platforms, or third-party hosting services
Create applications or services that provide access to the dataset for others

2. Access, Sharing, and Distribution

a. Access Requirements

Access is granted only upon:

Submission of a signed Data Use Agreement
Verification and approval by the CHiME-9 ECHI Challenge Organizers

b. Email address and affiliation

Where possible, please request access using an email address associated with your institution, employer, or research organisation. Requests made using personal email addresses such as Gmail, Outlook, Yahoo, or similar may take longer to process, as we may need additional information to verify affiliation and intended use.

If you do not have an institutional or work email address, please explain this in the intended-use field and provide sufficient information for us to assess the request.

c. Redistribution Policy

Redistribution of the dataset or identifiable components to third parties is strictly prohibited. This includes:

Raw data (e.g., audio, tracking data, metadata)
Reversible or personally identifiable processed data (e.g., biometric embeddings)

d. Sharing of Derived Works

You may share trained models or results only if:

They do not expose raw or identifiable data
They are not reversible to reconstruct original data
You provide proper attribution as described in Section 3

3. Attribution and Citation

If you use the CHiME-9 ECHI Dataset in any publication, presentation, or shared work, you must include the following citation:

Sutherland, R., Clarke, J., Elghazaly, H., Kuebert, T., Lugger, M., Petrausch, S., Azcarreta Ortiz, J., Xu, B., Goetze, S. and Barker, J., "Descriptor: Enhancing Conversations for the Hearing Impaired in the 9th Computational Hearing in Multisource Environments Challenge (CHiME9 ECHI)," in IEEE Data Descriptions, vol. 3, pp. 73-81, 2026, doi: 10.1109/IEEEDATA.2025.3647612.

You must also:

Link to the dataset’s official access page
Acknowledge the CHiME-9 ECHI Challenge Organizers as the data providers

4. Privacy, Ethics, and Compliance

You agree to:

Not attempt to re-identify individuals
Not use the dataset for biometric surveillance, profiling, or other unethical purposes
Comply with all relevant privacy and data protection laws, including the General Data Protection Regulation (GDPR) where applicable
Conduct research using this dataset in accordance with your institution's ethical standards or Institutional Review Board (IRB) requirements

5. Data Security

You agree to:

Store the dataset in secure, access-controlled environments
Prevent unauthorized access, sharing, or duplication
Refrain from uploading the dataset to public cloud platforms or storage services (e.g., Google Drive, Dropbox, GitHub)

6. Termination and Enforcement

Any violation of this agreement may result in immediate termination of access
Upon request or termination, you must permanently delete all copies of the dataset
You must report any known or suspected misuse within 48 hours to:
echichallenge@gmail.com

7. Disclaimer of Warranty

The CHiME-9 ECHI Dataset is provided "as is", without warranties of any kind. The CHiME-9 ECHI Organizers expressly disclaim:

Any warranties of title or non-infringement
Implied warranties of merchantability or fitness for a particular purpose
Any guarantee of dataset accuracy, quality, or usefulness

8. Limitation of Liability

To the fullest extent allowed by law, the CHiME-9 ECHI Challenge Organizers shall not be liable for any:

Indirect, incidental, or consequential damages
Losses related to use or inability to use the dataset
Claims resulting from third-party use or interpretation of your derivative works

9. Contact and Licensing Requests

For questions regarding this agreement or to inquire about commercial licensing, please contact:
📧 echichallenge@gmail.com

If you use this dataset, please cite:

Sutherland, R., Clarke, J., Elghazaly, H., Kuebert, T., Lugger, M., Petrausch, S., Azcarreta Ortiz, J., Xu, B., Goetze, S. and Barker, J., "Descriptor: Enhancing Conversations for the Hearing Impaired in the 9th Computational Hearing in Multisource Environments Challenge (CHiME9 ECHI)," in IEEE Data Descriptions, vol. 3, pp. 73-81, 2026, doi: 10.1109/IEEEDATA.2025.3647612.

BibTeX

@ARTICLE{11314168,
  author={Sutherland, Robert and Clarke, Jason and Elghazaly, Hend and Kuebert, Thomas and Lugger, Marko and Petrausch, Stefan and Ortiz, Juan Azcarreta and Xu, Buye and Goetze, Stefan and Barker, Jon},
  journal={IEEE Data Descriptions}, 
  title={Descriptor: Enhancing Conversations for the Hearing Impaired in the 9th Computational Hearing in Multisource Environments Challenge (CHiME9 ECHI)}, 
  year={2026},
  volume={3},
  number={},
  pages={73-81},
  keywords={Oral communication;Tracking;Microphones;Recording;Loudspeakers;Auditory system;Speech enhancement;Noise measurement;Background noise;Glass;Computational Hearing in Multisource Environments (CHiME);conversational speech;speech enhancement (SE);speech in noise;target speaker extraction},
  doi={10.1109/IEEEDATA.2025.3647612}}

The main CHiME-9 ECHI dataset consists of recordings of conversations between four participants seated around a table. During the sessions, a cafeteria environment is simulated using a set of 18 loudspeakers positioned in the recording room. In each session one participant is using Aria glasses (Project Aria) and another is wearing hearing aid devices with two microphones per ear. The task is to enhance the audio arriving at the device microphones to improve it for the purposes of conversation.

The figure below shows a photo of the recording room showing four participants engaged in the task (left); and a schematic illustrating how the loudspeakers are used (right).

Additional details:

14 of the loudspeakers are used to simulate up to 7 simultaneous two-person conversations, using data from either the LibriSpeech data or the spontaneous speech samples from the EARS datasets.
Additional sound events from FSD50K were sparsely mixed into the 14 conversation channel. These were selected using the semantic tags using sounds considered appropriate for a cafeteria environment, e.g. clinking glasses, coughs, etc.
Ambient sound was added using four additional loudspeakers placed facing the room corners and playing background audio from the WHAM! dataset.
All recordings were made in the same room but variability was introduced by making small movements to the position and direction of the 14 conversational loudspeakers, and my changing the configuration of the room's absorber panels between sessions.

For instructions for obtaining the data jump to Getting the data.

Description of the ECHI dataset

The data has been provided from 49 separate recording sessions, each of 36 minutes duration. These have been split into training, development and evaluation subsets. Each dataset is fully disjoint with respect to the conversation participants, and the LibriSpeech/EARS speakers, the ambient noise environments and the sound effect materials that are used to create the interfering sound during the sessions. The evaluation subset will remain hidden until shortly before the final submission of the system. A baseline system is provided that has been trained on the training subset and evaluated on the development subset. Participants can use the training subset to build their system and the development subset to compare against the baseline. No training or automatic fine-tuning is allowed on the development set.

split	background	# sessions	# speakers	duration
train	LibriSpeech	30	118	18.0 h
dev	LibriSpeech	4	16	2.4 h
	EARS	6	24	3.6 h
eval	LibriSpeech	5	20	3.0 h
	EARS	4	16	2.4 h

The development and evaluation subsets each contain sessions using both LibriSpeech and EARS spontaneous speech as background sources. These are treated as unified sets—evaluation scores will be averaged across all sessions within each subset, irrespective of background type.

For each session, the following are provided

audio recorded by the Aria glasses (7 channels)
audio recorded by the Hearing Aids (4 channels)
close talk audio (1 channel per participant)
head position and orientation (per participant), 250 Hz frame-rate
participant speech (clean, read speech recordings for target speaker extraction)

All audio is recorded at 48 kHz.

Note above that the locations at the table are numbered 1, 2, 3 and 4. These position numbers are used in the metadata, e.g. to identify the location of the person wearing the Aria glasses and the hearing aid, and to associate close talk microphone and head tracking data with participants.

Detailed description of data structure and formats

The directory structure is as follows (note that the evaluation subdirectories will be distributed shortly before the submission deadline):

├── aria
│   ├── train
│   ├── dev
│   └── eval
├── ha
│   ├── train
│   ├── dev
│   └── eval
├── ct
│   ├── train
│   ├── dev
│   └── eval
├── ref
│   ├── train
│   ├── dev
│   └── eval
├── tracker
│   ├── train
│   ├── dev
│   └── eval
├── participant
│   ├── train
│   ├── dev
│   └── eval
└─── metadata
    ├── train/ref
    ├── dev/ref
    └── eval/ref

Formats of the data in the subdirectories are as follows:

aria contains WAV files with 7-channel recordings, with a 48 kHz sampling rate. The stem of the audio file name will be used as ID of the recording in all the subsequent filenames and lists (<session-id>.aria.wav).
ha contains WAV files with 4-channel recordings, with a 48 kHz sampling rate, named <session-id>.ha.wav. Channels 1 and 2 are the front microphones of the left and right hearing aid; channels 3 and 4 are the back microphones of the left and right hearing aids.
ct contains the close-talk mic WAV files with 48 kHz sampling rate. There are four single-channel signals <session-id>.ct_pos1.wav, <session-id>.ct_pos2.wav, <session-id>.ct_pos3.wav and <session-id>.ct_pos4.wav. The position numbers correspond to the wearer's location at the table. The position of the hearing aid and Aria wearer is available in the metadata.
ref contains the reference signals for computing intrusive metrics. Separate references are provided for the Aria and the hearing-aid. <session-id>.aria.pos1.wav, ..., <session-id>.aria.pos4.wav, <session-id>.ha.pos1.wav, ..., <session-id>.ha.pos4.wav, with pos indicating the seating position. For convenience, symbolic links to the same files are provided named with the correspoinding participant ID, i.e. <session-id>ha.P001.wav etc. The reference is a single-channel audio file with the duration of the session. The metadata contains a csv file identifying the valid speech regions within each reference file.
tracker contains the head Vicon tracking data as a separate csv file for each participant: <session-id>.tracker_pos1.csv, <session-id>.tracker_pos2.csv, <session-id>.tracker_pos3.csv and <session-id>.tracker_pos4.csv. The position index has the same meaning as for the close-talk mic data. Within each csv file, there are six column, TX, TY, TZ, RX, RY and RZ, giving the locations and orientations of the heads at a 250 Hz framerate. The data is in mm and radians using a helical angle convention.
participant contains WAV files for each participant storing, the participant's reading of the 'Rainbow passage' that can be used for obtaining a speaker embedding. The files are named <participant-id>.wav, where the metadata records the participant-id of each participant in each recording position in each session.
metadata - Metadata for each session, identifying participants in each session, location of the Aria glasses and Hearing aids.
metadata/ref - Contains csv files that give the start and end sample of each valid speech region in the reference file. These are the regions over which speech quality metrics will be computed. The files are named <session-id>.<device>.<position>.csv. For convenience, symbolic links to the same data are provided with the names <session-id>.<device>.<participant-id>.csv where paritcipant-id is the id of the participant sitting at the position, and correspond to the IDs used for the participant WAV files.

During test time, systems can only make use of the aria, ha and participant data plus the metadata that identifies the participants in the session. The ct and tracker data can only be used during training, and the ref and metadata\ref data can only be used for analysis or evaluation purposes.

Metadata files

In addition to the reference csv files, the metadata directory also contains session csv files, session.<datasplit>.csv. These provide the participant ID of the participant at each seating position, and also indicate which seat position was wearing the hearing aids and aria glasses. These have the following format,

session,aria_pos,ha_pos,pos1,pos2,pos3,pos4
train_01,3,4,P005,P007,P008,P006
train_02,4,3,P011,P010,P009,P012
train_03,1,2,P014,P016,P015,P013
train_04,1,2,P029,P030,P031,P032
...

So for example, for session train_01, the Aria glasses are worn by the participant in position 3 (P008), and the hearing aids by the participant at position 4 (P006). Note, the Aria glasses were not available for three of the training sessions: train_16, train_28 and train_29 . This is indicated in the metadata file by the position number being omitted from the aria_pos column for these rows.

Hearing aid devices

The hearing aid (HA) devices have two microphones (front and back) for the left and right ear, to give a total of four channels. The hearing aid signals are preamplified and then connected to the same audio interface as the close talk (CT) microphones allowing for sample alignment across all HA and CT channels.

The figure below shows the same device from two different orientations with the approximate locations of the front and back microphone indicated.

Aria Glasses

Detailed information on the Aria glasses can be found at the website of the Aria project, especially in the documentation.

Two sizes of Aria Glasses were used 'Big' and 'Small' chosen according to which fit the participant most comfortably. The microphone geometry for the Large glasses is as follows.

position	x [cm]	y [cm]	z [cm]
	back → front	right → left	bottom → up
lower-lens right	9.95	-4.76	0.68
nose bridge	10.59	0.74	5.07
lower-lens left	9.95	4.49	0.76
front left	9.28	6.41	5.12
front right	9.93	-5.66	5.22
rear right	-0.42	-8.45	3.35
rear left	-0.48	7.75	3.49

The Aria audio signal are recorded to storage on the glasses. After each session these signals are recovered. Note, there is a clock-drift between the Aria glasses and MOTU audio interface used to record all other audio signals. This drift is at the rate of about 100 ms per hour and has been carefully estimated per session to allow alignment of signals (e.g. see Reference Signal construction below).

Reference Signals

Reference signals are provided for training enhancement systems and evaluating performance. These signals approximate the clean, direct-path speech from conversation participants as received at the device microphones. Since this is real-world data (not simulated), references are constructed using close-talk microphones with the following process:

Construction Process:

Close-Talk Processing
- Directional close-talk mics are denoised to reduce background leakage
- High SNR is maintained before/after processing
Speech Segmentation
- SileroVAD identifies active speech regions
Delay & Synchronization
- Propagation delay: Calculated using head-tracker distance measurements between speaker and listener
- Clock drift correction (Aria only): Linear compensation applied between Aria/close-talk clocks
Fine Alignment
- ±2ms adjustment via cross-correlation with device mics:
  - Aria: Nose bridge microphone
  - Hearing aids: Computed for front left/right mics then averaged
Reference Assembly
- Time-aligned segments are summed to create single-channel references

Key Characteristics:

📌 Format: Single-channel (despite multi-channel devices)
⚠️ Limitations:
- No microphone/HRTF characteristic compensation
- Quality depends on close-talk denoising performance
- Spatial effects not captured
- Time-delays modelled using average over the speech segment
✅ Suitable for:
- Training objectives
- STOI and similar metrics
- STFT-based loss functions
❌ Not suitable for:
- Metrics requiring exact multi-channel reproduction
- Binaural/HRTF evaluation

Tracker Data

The head tracking data is stored in csv files named <session-id>.tracker_<pos>.csv. Where pos can be pos1, pos2, pos3 or pos4 and indicates the seat of the participant being tracked. The csv file has the following format:

RX, RY, RZ, TX, TY, TZ
0.113541,0.037043,2.332312,299.732452,-155.289459,1320.493652
0.114847,0.035577,2.331669,299.915100,-155.544464,1320.511963
0.115966,0.033263,2.329266,300.026062,-155.711182,1320.475342
0.118553,0.032524,2.330051,300.332703,-156.053162,1320.554688
0.120550,0.031226,2.329265,300.456024,-156.267731,1320.530640
0.122704,0.029718,2.328243,300.627594,-156.503647,1320.539795
...

The tracking data provides the position and orientation of the tracking hat that is being worn by the participant. Each row represents a successive frame at a frame rate of 250 Hz. The data has been carefully synchronised with the start of the audio recordings. The RX, RY and RZ present the orientation as a rotation stored using a helical format. i.e. the vector RX, RY, RZ is in the direction of the axis of the rotation, and the magnitude of the vector equals the angle of the rotation measured in radians. The TX, TY and TZ give the world coordinations of the tracking objects 'origin' measured in millimeters.

The geometry of the object being tracked is stored in Vicon vsk files in the metadata directory. These files specify the object's coordinate system with its origin and the location of its constituent markers.

Using the tracking data and the vsk file, the trajectory of the individual markers can be reconstructed. The figure below shows a 5-second animation of typical participant head motion from a recording session. This has been reconstructed from the tracker csv file and the 'vsk' files which store the relative marker positions in each object being tracked.

Getting the data

CHiME-9 ECHI Dev:
- Includes dev data signals for HA and Aria plus essential metadata
- chime9_echi_dev.tar.gz [~23.6GB]
CHiME-9 ECHI Train:
- Includes all the training data signals for HA and Aria.
- chime9_echi.train_pt1.v1_0.tar.gz [~30GB]
- chime9_echi.train_pt2.v1_0.tar.gz [~37GB]

(The training data has been split into two parts to keep within HuggingFace's 50 GB max file size. Please download both).

For obtaining the the CHiME-9 ECHI dataset, you first need to request access and then download it using your Hugging Face token. Please note that access requires signing a Data Use Agreement.

Request Access:
- Go to the CHiME-9 ECHI dataset repository on Hugging Face.
- Request access to the dataset. You will need a Hugging Face account.
Get Your Hugging Face Token:

If you don't have one, create a Hugging Face User Access Token. Follow the instructions here to obtain your HF_TOKEN. Make sure it has read permissions.
Download and Untar Dataset Files:

It's recommended to download and untar the data into a data directory within your CHiME9-ECHI project folder. (See baseline.md.)

# Navigate to your baseline project directory if you aren't already there
# cd CHiME9-ECHI

# Create a data directory (if it doesn't exist)
mkdir -p data
cd data

# Export your Hugging Face token (replace 'your_actual_token_here' with your actual token)
export HF_TOKEN=your_actual_token_here

# Download the development set
wget --header="Authorization: Bearer $HF_TOKEN" https://huggingface.co/datasets/CHiME9-ECHI/CHiME9-ECHI/resolve/main/data/chime9_echi.dev.v1_0.tar.gz

# Untar the downloaded files 
tar -zxovf chime9_echi.dev.v1_0.tar.gz

The training data is split into two parts and can be downloaded similarly

wget --header="Authorization: Bearer $HF_TOKEN" https://huggingface.co/datasets/CHiME9-ECHI/CHiME9-ECHI/resolve/main/data/chime9_echi.train_pt1.v1_0.tar.gz
wget --header="Authorization: Bearer $HF_TOKEN" https://huggingface.co/datasets/CHiME9-ECHI/CHiME9-ECHI/resolve/main/data/chime9_echi.train_pt2.v1_0.tar.gz

# Untar the downloaded files 
tar -zxovf chime9_echi.train_pt1.v1_0.tar.gz
tar -zxovf chime9_echi.train_pt2.v1_0.tar.gz

Untar all files in the same directory. The '-o' flag is needed because some small files, e.g. the metadata, are stored redundantly in each tar file for convenience, so they should be allowed to overwrite when unpacking.

License and Data Use Agreement

Dataset: CHiME-9 ECHI (Enhancing Conversations to address Hearing Impairment) Dataset
Version: June 2025
Maintained by: CHiME-9 ECHI Challenge Organizers
Contact: echichallenge@gmail.com

This dataset contains sensitive audio recordings, including identifiable speech of multiple speakers.

Use of this dataset is restricted under the terms of the CHiME-9 ECHI Data Use Agreement.

✅ Permitted Uses

Academic and educational purposes
Algorithm development and benchmarking
Publication of models and results with attribution

❌ Prohibited Uses

Redistribution of raw data
Uploading the dataset to public cloud or sharing platforms
Biometric surveillance, re-identification, or profiling

📚 Citation

If you use this dataset, please cite:

Sutherland, R., Clarke, J., Elghazaly, H., Kuebert, T., Lugger, M., Petrausch, S., Azcarreta Ortiz, J., Xu, B., Goetze, S. and Barker, J., "Descriptor: Enhancing Conversations for the Hearing Impaired in the 9th Computational Hearing in Multisource Environments Challenge (CHiME9 ECHI)," in IEEE Data Descriptions, vol. 3, pp. 73-81, 2026, doi: 10.1109/IEEEDATA.2025.3647612.

BibTeX

@ARTICLE{11314168,
  author={Sutherland, Robert and Clarke, Jason and Elghazaly, Hend and Kuebert, Thomas and Lugger, Marko and Petrausch, Stefan and Ortiz, Juan Azcarreta and Xu, Buye and Goetze, Stefan and Barker, Jon},
  journal={IEEE Data Descriptions}, 
  title={Descriptor: Enhancing Conversations for the Hearing Impaired in the 9th Computational Hearing in Multisource Environments Challenge (CHiME9 ECHI)}, 
  year={2026},
  volume={3},
  number={},
  pages={73-81},
  keywords={Oral communication;Tracking;Microphones;Recording;Loudspeakers;Auditory system;Speech enhancement;Noise measurement;Background noise;Glass;Computational Hearing in Multisource Environments (CHiME);conversational speech;speech enhancement (SE);speech in noise;target speaker extraction},
  doi={10.1109/IEEEDATA.2025.3647612}}

⚠️ Legal & Ethical Requirements

Comply with applicable laws including the GDPR (EU data protection law)
Follow your institution’s research ethics process (e.g., IRB review)
Do not attempt to re-identify individuals
Store data securely in access-controlled environments

Downloads last month: 33

Total file size:

105 GB