Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
SpeeChin (often misspelled “Speechin”) was a real Cornell-led research prototype: a necklace with an infrared camera that looked up at the wearer’s chin and lower face to classify a limited set of silently mouthed commands. It did not read thoughts or transcribe arbitrary silent speech, and the research does not establish that it became a product people can buy.
What SpeeChin was designed to do
Voice assistants generally need a microphone to pick up audible speech. That can be awkward in a library, meeting, shared office, or other quiet setting; noise can also make voice input harder. SpeeChin explored a different approach: infer a command from movements the wearer makes while mouthing it, without relying on the sound of a spoken voice.
The researchers also noted possible relevance for people unable to produce ordinary vocal speech. That is a potential application, not evidence that SpeeChin was clinically tested, approved, or shown to work as an assistive communication device.
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →The project’s official name is SpeeChin: A Smart Necklace for Silent Speech Recognition. Its paper appeared in the December 2021 issue of Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies (volume 5, issue 4, article 192; DOI 10.1145/3494987). Cornell published a project summary in February 2022.
#1 Best Overall
- Battery and Memory： Powered by a built-in rechargeable lithium battery, this small voice recorder supports up to 24 hours of continuous recording on a single charge. The 8GB memory allows you to store up to 100 hours of audio at the default transmission rate of 256 kbps, or up to 500 hours at 32 kbps
- Portable Necklace Pendant Design：Tiny in size. Works not only as a wearing recording device for daily documentation, but also as an MP3 music player. Ideal for classes, meetings, interviews and more occasions.
- High‑Fidelity Recording with Noise Reduction： Supports 1536KBPS high bit‑rate recording for true‑to‑life sound reproduction. Built‑in smart digital noise reduction cuts background interference, producing clearer audio.
- Easy Operation: The mini recording device features simple one-button operation. Just push the on/off button to another side to record. Push it back to switch off the device
- Plug & Play Anywhere：Comes with a earphone for instant audio playback anytime, anywhere. Just connect the earphone to the device and then push the switch on the power button to play. press + on the earphone to skip between audio files, or long press + to adjust the volume
How the necklace detected a command
SpeeChin used a customized infrared camera mounted in a 3D-printed necklace housing, suspended from a silver chain. Stabilizing side wings and a weight at the bottom helped keep the camera in position. Rather than facing outward at the wearer, the camera pointed upward from below the chin toward the lower face and neck.
- The wearer silently mouthed or articulated a command from the system’s known vocabulary.
- Movement of the tongue, jaw, lips, and facial muscles changed the shape and appearance of the skin beneath the chin and around the neck.
- The infrared camera captured those visual changes. The researchers’ central insight was that useful cues might be visible from below, even without a frontal view of the lips.
- Image processing prepared the camera data, and a deep convolutional-recurrent neural network (CRNN) classified the utterance.
- The resulting command could be sent to a paired phone or another computing device.
This is silent speech recognition (SSR): a system infers an utterance from physical articulation or related signals without depending on audible speech. It is different from ordinary voice recognition, which uses sound, and very different from thought decoding. If a wearer merely thinks a command without making the movements the camera can observe, SpeeChin has no basis to recognize it.
What it recognized—and what it did not
The main evaluation used a fixed, or closed-set, vocabulary: 54 English commands and 44 Chinese commands. The English set included categories such as digits, interactive and voice-assistant commands, punctuation, and navigation commands. The Chinese set contained selected words and phrases. The classifier chose among known options; the study did not show that the necklace could understand any sentence a wearer might silently say or provide unrestricted dictation.
Rank #2
- Easy to Operate: The mini recorder supports simple one click and voice control, supports fast forward and fast rewind functions, and also supports time setting.
- Voice Activation: Automatically start recording when noise is detected, and the recording is within five meters, freeing up your hands so you can focus on other tasks.
- Long Battery Life and Memory: Built in rechargeable lithium battery, audio recorder can record up to 20 hours continuously after full , and can be used for about 15‑20 days, and with up to 100 hours of audio file storage memory.
- Compact and Portable: The recorder is small in size, easy to carry and hide, not only a portable recorder to record anything, or an MP3 player to listen to your favorite songs, and can be widely used in classroom, conference, interview recording and so on.
- Noise Reduction: The recording bit rate can reach 192KBPS, the sound is more real, and the built in intelligent digital noise reduction technology can better the environmental noise, and the recording content is clearer.
The researchers also tested 72 one-syllable nonwords, constructed from phoneme combinations, with 10 participants. This exploratory work investigated recognition of more basic speech units and the possibility of expanding vocabularies. It should not be mistaken for a demonstration of general-purpose silent transcription.
Accuracy: strong in controlled tests, lower while walking
In the main, controlled studies, 20 participants were split between English and Chinese evaluations, with 10 participants in each language group. The reported average cross-session accuracy was:
| Language | Commands tested | Participants | Average cross-session accuracy |
|---|---|---|---|
| English | 54 | 10 | 90.5% |
| Chinese | 44 | 10 | 91.6% |
Cross-session evaluation matters because the researchers assessed performance after the necklace was removed and remounted between sessions, rather than relying only on one continuous recording setup. But the figures still describe recognition among a particular set of commands, in a small study—not the chance of correctly understanding any silent sentence in everyday use.
Rank #3
- Elegant Heart-Shaped Design: This discreet voice recorder combines functionality with style, featuring a heart-shaped design that serves as a fashionable recording necklace. Perfect for everyday carry, it allows you to capture audio without drawing attention.
- Advanced AI Voice-to-Text Technology: Equipped with cutting-edge AI, this recording necklace converts your spoken words into text with impressive accuracy. It automatically activates when sounds are detected within a 5-meter range, ensuring you never miss important discussions in work or interviews.
- Impressive 200 Hours of Storage: Enjoy the luxury of up to 200 hours of recording time with our discreet voice recorder. With a high-quality 192KBPS bit rate and advanced HD noise reduction, it captures clear audio even in challenging environments like meetings or busy lectures.
- User-Friendly One-Click Control: Experience effortless operation with the one-click control feature for recording and playback. The auto-save function ensures that your recordings are preserved, even if the device unexpectedly powers off, safeguarding your valuable notes.
- Portable and Versatile Companion: Designed for convenience, this recording necklace includes a discreet lanyard hole for easy attachment to keys or wearing as jewelry. Ideal for journalists, students, and professionals, it’s the perfect portable solution for reliable recording on the go.
Accuracy rose when the choice set was narrower. A 21-command English subset of digits and punctuation averaged 93.5%; a 19-phrase Chinese subset averaged 94.8%. That pattern is consistent with the system having fewer possible commands to distinguish.
Recommended Free Tools
Movement exposed a more practical limitation. In a small walking study of six participants, average phrase accuracy fell to 72.3% for English and 65.5% for Chinese. Results varied markedly by person: English scores ranged from 40.6% to 91.9%, and Chinese from 34.4% to 91.2%. The researchers attributed much of the drop to head movement and differences in walking style. Thus, the headline 90%-plus results do not describe performance on the move.
Prototype hardware, power, and processing
The prototype’s infrared imaging system was built to help separate skin from background. The researchers had considered a standard RGB camera, but found background segmentation difficult; thermal and depth cameras were considered too large or unsuitable for the intended form factor. The selected camera system used two customized infrared LED boards with 850-nanometer LEDs, an OV5647 image sensor, a 320 × 240 image stream at 60 frames per second, a 130-degree field of view, and a dual-peak 650+850-nanometer narrow-band optical filter. These are research-prototype specifications, not specifications for a retail device.
Rank #4
- 【Compact & Portable for Anytime, Anywhere Use】 - This compact, slim-profile voice recorder supports versatile carrying options: wristband, lanyard, keychain, or magnetic attachment. Wear it like a watch on your wrist, drape it like jewelry across your chest, or clip it to your keys. Its magnetic feature securely attaches to iron surfaces like notebooks, refrigerators, or computer cases.Compact and space-saving, it effortlessly accompanies you during business commutes, study outings, or daily travels. Combines practicality with stylish decor.
- 【3072Kbps HD Audio Quality】This digital voice recorder enhances environmental sound recognition and optimizes vocal clarity, achieving recording quality up to 3072Kbps. It restores the original texture of voices, instruments, and other sounds, delivering CD-quality audio. Whether for business negotiations or music creation, it preserves sound without any loss.
- 【DSP 3.0 Noise Reduction & HD Recording】Features a built-in professional DSP acoustic processor with adjustable noise reduction levels (1-7). It intelligently filters ambient noise in noisy environments—be it bustling conference rooms, crowded streets, or classrooms with background chatter—to focus on target sounds, ensuring crystal-clear recordings.
- 【One-Touch Recording】Simplifies operation with no complex settings required. Pressing the button simultaneously powers on the device and starts recording. Responds instantly to sudden inspiration, impromptu conversations, or brief yet crucial information, ensuring no valuable audio moment is missed.
- 【Supports Wired Headphone Controls】Plays MP3, WAV, APE, FLAC, WMA, and more. Plug in headphones during breaks to unwind with music, or use it for language learning during commutes—one device, dual purposes.
The sensing hardware consumed about 5.4 watts, excluding the data-processing pipeline running on a remote machine. The paper breaks that estimate into 1.48 W for the two LEDs, 0.92 W for the camera module, and 3 W for a Raspberry Pi 4B. The deep-learning pipeline relied on remote computing in the reported prototype configuration, so standalone, phone-free operation should not be assumed. The authors described the system as a proof of concept and suggested lower-power processors and cameras, along with reduced LED duty cycles, as possible ways to improve it. The published results do not establish smartwatch-like battery life.
Privacy: a different camera angle, not camera-free sensing
A camera aimed mainly at the underside of the chin, ceiling, or sky may expose less of a wearer’s surroundings than a camera pointed directly at the face. That was a privacy advantage the design sought, not a guarantee of privacy. The device still used a camera and could capture environmental information. The paper discusses an activate-to-use approach and future options such as on-device feature extraction and removing privacy-sensitive information from images; these are design considerations, not proof of a consumer privacy certification or comprehensive data-handling policy.
Conditions that could undermine recognition
- Walking, turning, or shifting the necklace: Movement changes the camera’s view and can reduce accuracy. The walking results show substantial performance loss and person-to-person variation.
- Hair or clothing over the camera: Long hair can cover the lens, while clothing can obstruct it or push the necklace out of position.
- Strong sunlight: Direct sunlight can make it harder to distinguish skin from the background in infrared imagery.
- Untrained words or sentences: The main system classified fixed commands. The nonword experiment did not demonstrate arbitrary silent dictation.
- Different articulation, fluency, or language background: The study was small. Chinese participants were native speakers, while the English participants were not native English speakers; the authors noted that fluency and pronunciation could affect results.
- Continuous use: Power demand and camera-related privacy questions become more significant if sensing is always on. The paper discusses activation when needed as a possible mitigation.
These constraints help explain why the paper also discusses possible redesigns, including a different form factor or tie clip, correcting for body position in post-processing, and collecting training data with more head movement. They are future engineering directions, not features demonstrated as solved in the prototype.
Best Value
- 【Easy Portability for On-the-Go Use】- This voice recorder portable is designed for effortless carrying, fitting easily into bags or pockets for use anywhere. The voice activated recorder responds quickly to your needs, while its sturdy build protects it from daily wear and tear. A reliable companion for recording anytime, anywhere.
- 【DSP 6.0 Noise Reduction & HD Recording】- This sound recorder adopts advanced DSP 6.0 technology to filter out background noise, ensuring crisp audio even in busy environments. The high-definition capture of the audio recorder preserves every detail of conversations, while the voice recorder delivers natural, distortion-free sound for important content. A reliable choice for clear recordings anytime.
- 【OTG Function for Easy Data Transfer】- The integrated OTG feature of this audio recording device allows direct connection to smartphones or tablets, enabling instant data transfer without a computer. This recorder device simplifies saving or editing recordings on mobile devices, fitting seamlessly into your daily digital routine for maximum convenience.
- 【72GB Large Storage & 10860H Capacity】- With 72GB built-in memory, this recorder can store up to 10860 hours of audio files, eliminating frequent file deletion. The voice recorder with playback lets you access recordings easily, while its efficient storage management supports long-term use without hassle.
- 【Versatile Uses for Various Scenarios】Ideal as a recorder for lectures, this voice recorder also excels in business meetings, interviews, and personal memos. It adapts to both indoor and outdoor environments, delivering consistent performance wherever you need to capture audio. A multi-purpose tool for all your recording needs.
Can you buy a SpeeChin necklace?
The project page, Cornell’s coverage, and the research paper establish SpeeChin as an academic prototype, but do not establish retail availability, a preorder, price, consumer setup process, or product-support channel. As of the sources documented through August 18, 2026, it is safest to treat SpeeChin as research rather than a purchasable consumer wearable. That is not a claim that commercialization is impossible; it is a statement that the cited project materials do not show that it happened.
Why the research still matters
SpeeChin demonstrated a distinctive wearable-input idea: a necklace can use an upward-facing infrared camera to pick up changes around the chin and neck, rather than asking the user to speak aloud or placing a camera in front of the face. Its evaluation across sessions, in English and Chinese, and under walking conditions also made visible both the promise and the gap between a controlled lab result and reliable everyday use.
The accurate takeaway is narrower than “a necklace that understands silent speech”: SpeeChin classified selected, silently mouthed commands with promising accuracy under controlled conditions. Movement, vocabulary limits, power, environmental factors, small samples, and the absence of demonstrated commercial availability all matter when judging what it could do.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsSources: The full research paper; Cornell’s project summary; SpeeChin project page.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
