1. Gabatarwa
Bukatar hulɗar Mutum da Kwamfuta (HCI) ta halitta kuma mai hankali tana ƙaruwa cikin sauri, wanda aikace-aikace a cikin wasanni, gidaje masu wayo, da mu'amalar motoci ke tafiyar da shi. Duk da haka, hanyoyin hulɗa na al'ada suna fuskantar manyan iyakoki: allon tabawa ya kasa a cikin yanayi mai ɗanɗano/maiko, kyamarori suna tayar da damuwa game da sirri kuma suna da yawan amfani da wutar lantarki, kuma sarrafa murya yana fama da umarni masu rikitarwa da al'amuran sirri. Kasuwar HMI ta duniya ana hasashen za ta kai dalar Amurka biliyan 7.24 nan da shekara ta 2026, wanda ke nuna buƙatar gaggawa don mafi kyawun mafita.
Wannan takarda ta gabatar da EMGesture, sabuwar fasahar hulɗa ba tare da taɓawa ba wacce ke mayar da cajin Qi mara waya da ake samu ko'ina a matsayin na'urar gano motsin hannu. Ta hanyar nazarin siginonin lantarki (EM) da ake fitarwa yayin caji, EMGesture tana fassara motsin hannun mai amfani ba tare da buƙatar ƙarin kayan aikin hardware ba, tana magance tsada, sirri, da ƙalubalen gama gari da ke tattare da wasu hanyoyi.
97%+
Ingantaccen Gane
30
Mahalarta
10
Na'urorin Wayar Hannu
5
Cajin Qi da aka Gwada
2. Hanyoyi & Tsarin Tsarin
EMGesture ta kafa tsari mai zuwa daga farko zuwa ƙarshe don gane motsin hannu ta amfani da "gefen tashar" EM na cajin Qi.
2.1. Samun Sigina na EM & Gyara Kafin Aiki
Tsarin yana ɗaukar ainihin siginonin lantarki da coil ɗin watsa wutar lantarki a cikin cajin Qi ke samarwa. Wani mahimmin fahimta shine cewa motsin hannu kusa da cajin yana dagula wannan filin EM ta hanya mai aunawa kuma ta musamman. Ainihin siginar, $s(t)$, ana samfurinta sannan ta bi ta hanyar gyara kafin aiki:
- Tacewa: Tace mai tacewa yana cire hayaniyar mitar girma da karkatar da mitar ƙasa, yana ware mitar da ta dace da motsin hannu.
- Daidaituwa: Ana daidaita siginoni don yin la'akari da bambance-bambance a cikin samfuran caji da sanyawar na'ura: $s_{norm}(t) = \frac{s(t) - \mu}{\sigma}$.
- Rarraba: Ana sanya bayanan ci gaba zuwa sassa masu dacewa da abubuwan motsin hannu ɗaya ɗaya.
2.2. Cire Siffofi & Rarraba Motsin Hannu
Daga kowane sashe da aka riga aka gyara, ana cire tarin siffofi masu yawa don siffanta tasirin motsin hannu akan filin EM.
- Siffofi na Lokaci: Matsakaici, bambanci, ƙimar ketare sifili, da ƙarfin siginar.
- Siffofi na Yankin Mita: Cibiyar bakan, faɗin bakan, da ƙididdiga daga Canjin Fourier na Gajeren Lokaci (STFT).
- Siffofi na Lokaci-Mita: Siffofi da aka samo daga canjin wavelet don ɗaukar kaddarorin siginar marasa tsayawa.
Waɗannan siffofi sun zama vector mai girma mai girma $\mathbf{f}$ wanda ake ciyarwa cikin ingantaccen mai rarraba na injin koyo (misali, Na'urar Tallafawa Vector ko Daji na Bazuwar) wanda aka horar da shi don tsara siffofin vector zuwa takamaiman alamun motsin hannu $y$ (misali, goge hagu, goge dama, danna).
3. Sakamakon Gwaji & Kimantawa
3.1. Ingantaccen Gane & Aiki
A cikin gwaje-gwajen da aka sarrafa tare da mahalarta 30 suna yin tarin motsin hannu na gama gari (misali, goge, da'ira, danna) akan cajin Qi daban-daban 5 da na'urorin wayar hannu 10, EMGesture ta sami matsakaicin ingantaccen gane wanda ya wuce 97%. Tsarin ya nuna ƙarfi a cikin samfuran caji daban-daban da nau'ikan na'urori, wani muhimmin al'amari don tura ko'ina. Matrix ɗin ruɗani ya nuna ƙaramin kuskuren rarrabuwa tsakanin azuzuwan motsin hannu daban-daban.
Bayanin Chati (Tunani): Chatin sandar zai iya nuna inganci kowane nau'in motsin hannu (duk sama da 95%), kuma chatin layi zai nuna ƙarancin jinkirin tsarin, tare da gane daga farko zuwa ƙarshe yana faruwa a cikin ƴan millisekonds ɗari, wanda ya dace don hulɗar ainihin lokaci.
3.2. Nazarin Mai Amfani & Kimanta Amfanin
Wani ƙarin nazarin mai amfani ya kimanta ma'auni na zahiri. Mahalarta sun ƙididdige EMGesture sosai akan:
- Dacewa: Yin amfani da na'urar da ke akwai (caji) ya kawar da buƙatar sabon kayan aikin hardware.
- Amfanin: An ɗauki motsin hannu a matsayin mai hankali kuma mai sauƙin aiwatarwa.
- Fahimtar Sirri: Masu amfani sun bayyana matakan jin daɗi sosai idan aka kwatanta da tsarin tushen kyamara, saboda babu bayanan gani da ke ciki.
4. Bincike na Fasaha & Fahimta ta Asali
Fahimta ta Asali
EMGesture ba wani takarda ne kawai na gane motsin hannu ba; darasi ne a cikin sake amfani da kayayyakin more rayuwa. Marubutan sun gano dandamali na kayan aikin hardware na gama gari, daidaitacce—cajin Qi—kuma sun yi amfani da fitar da EM da ba a yi niyya ba a cikin tashar ganowa mai mahimmanci. Wannan ya wuce daga dakin gwaje-gwaje kuma kai tsaye zuwa dakunan falo da motocin miliyoyin, yana ƙetare shingen karɓar da ke addabar yawancin binciken HCI na sabon abu. Hanya ce mai aiki, kusan wayo, don lissafin kwamfuta ko'ina.
Kwararar Ma'ana
Ma'ana tana da ban sha'awa mai sauƙi: 1) Matsala: Hanyoyin HCI da ke akwai suna da aibi (sirri, tsada, muhalli). 2) Lura: Cajin Qi suna ko'ina kuma suna fitar da filayen EM masu ƙarfi, masu gyara. 3) Hasashe: Motsin hannu na iya daidaita wannan filin ta hanyar da za a iya rarrabawa. 4) Tabbatarwa: Ingantaccen bututun ML ya tabbatar da inganci >97%. Kyawun yana cikin tsallake matakin "gina sabon na'urar gano" gaba ɗaya, kamar yadda masu bincike suka sake amfani da siginonin Wi-Fi don ganowa (misali, ganewar Wi-Fi don gano wurin zama) amma tare da mafi inganci da tushen siginar mai ƙarfi.
Ƙarfi & Aibobi
Ƙarfi: Al'amarin sirri-ta-zane shine siffa mai kashewa a yanayin yau. Rashin tsada ba za a iya musantawa ba—sifili ƙarin kayan aikin hardware ga mai amfani na ƙarshe. Ingantaccen 97% yana da ban sha'awa ga tsarin irin na farko. Aibobi: Giwa a cikin daki shine kewayo da ƙamus na motsin hannu. Takardar tana nuna alamar iyakokin kusanci; wannan ba na'urar gano dakin gaba ɗaya ba ce kamar wasu tsarin tushen radar. Saitin motsin hannu yana da yuwuwar asali kuma an iyakance shi zuwa motsi na 2D kai tsaye sama da caji. Bugu da ƙari, aikin tsarin na iya raguwa tare da caji na lokaci guda na na'urori da yawa ko a cikin yanayi mai hayaniyar lantarki—ƙalubale na ainihin duniya wanda ba a magance shi gaba ɗaya ba.
Fahimta masu Aiki
Ga manajoji samfura a cikin gida mai wayo da mota: Gwada wannan yanzu. Haɗa SDKs na EMGesture cikin tsarin nishadi na gaba ko na'urorin dafa abinci masu wayo. Dawowar zuba jari (ROI) a bayyane yake—ingantaccen aiki ba tare da ƙarin farashin BoM ba. Ga masu bincike: Wannan yana buɗe sabon ƙaramin fanni. Bincika tsararrun caji masu yawa don ganowa na 3D, koyon tarayya don samfuran keɓance ba tare da bayanai sun bar na'urar ba, da haɗawa da wasu na'urori masu gano ƙarancin wutar lantarki (misali, makirufo don umarnin "EM + murya"). Aikin Yang et al. akan ganowa na tushen RF (ACM DL) yana ba da tushen fasaha mai dacewa don ci gaba da wannan tsari.
Bincike na Asali & Hangen Nesa
Muhimmancin EMGesture ya wuce ma'aunansa na fasaha. Yana wakiltar sauyin dabarun binciken HCI zuwa ga ganowa na dama—yin amfani da abubuwan more rayuwa da ke akwai don dalilai da ba a yi niyya ba amma masu mahimmanci. Wannan ya yi daidai da manyan yanayin lissafin kwamfuta ko'ina, kamar yadda ake gani a cikin ayyuka kamar CycleGAN don fassarar hoto zuwa hoto mara biyu, wanda ke amfani da yankunan bayanai da ke akwai don samar da sababbi ba tare da biyu kai tsaye ba. Hakazalika, EMGesture yana amfani da yankin EM na caji da ke akwai don sabon yankin ganowa.
Daga mahangar fasaha, zaɓin siginonin EM akan madadin kamar Wi-Fi (misali, ganewar Wi-Fi) ko ultrasound yana da hikima. Ma'aunin Qi yana aiki a takamaiman mitar (100-205 kHz don ainihin bayanin martabar wutar lantarki), yana ba da siginar mai ƙarfi, daidaitacce, kuma mai keɓancewa idan aka kwatanta da cunkoson band 2.4/5 GHz. Wannan yana iya ba da gudummawa ga babban inganci. Duk da haka, dogaro da koyon inji don rarrabawa, yayin da yake aiki, yana gabatar da wani abu na "akwatin baƙi". Aikin nan gaba zai iya amfana daga haɗa ƙarin dabarun AI masu bayyanawa ko haɓaka samfuran zahiri waɗanda ke haɗa motsin motsin hannu kai tsaye zuwa ɓarnawar filin EM, kamar yadda aka bincika a cikin wallafe-wallafen ganowa na EM na asali da za a iya samun dama ta IEEE Xplore.
Da'awar inganci ta 97% tana da ban sha'awa, amma yana da mahimmanci a sanya shi cikin mahallin. Wannan yana yiwuwa inganci a cikin ƙayyadaddun yanayi na dakin gwaje-gwaje tare da ƙayyadaddun saitin motsin hannu. Tura ainihin duniya zai fuskanci ƙalubale kamar bambance-bambancen girman hannu, bambance-bambancen al'adu a cikin aiwatar da motsin hannu, da kutsawar lantarki na muhalli. Ƙarfin tsarin a kan waɗannan abubuwan zai zama gwaji na gaske na yuwuwar sa, ƙalubale gama gari ga yawancin tsarin ganowa kamar yadda aka lura a cikin kimantawa daga cibiyoyi kamar Cibiyar Ƙididdiga da Fasaha ta Ƙasa (NIST).
Misalin Tsarin Bincike
Yanayi: Kimanta EMGesture don sarrafa famfo na kicin mai wayo.
Aikace-aikacen Tsarin:
- Yiwuwar Sigina: Shin wurin caji (misali, tebur) ya dace don motsin hannu kusa da famfo? (Ee, mai yiwuwa).
- Taswirar Motsin Hannu: Taswirar motsin hannu masu hankali zuwa ayyuka: Goge hagu/dama don zafin jiki, motsin da'ira don sarrafa kwarara, danna don kunna/kashe.
- Binciken Ƙarfi: Gano hanyoyin gazawa: Fantsama ruwa (ba matsala ba ga EM), hannaye jika (babu matsala idan aka kwatanta da allon tabawa), tukwane na ƙarfe kusa (yuwuwar kutsawar EM—yana buƙatar gwaji).
- Tafiya na Mai Amfani: Mai amfani da hannaye masu maiko yana daidaita zafin ruwa ta hanyar goge akan kushin caji, ba tare da taɓa kowane sarrafawa na zahiri ba.
Wannan binciken na lamarin wanda ba na lamba ba yana kwatanta yadda ake kimanta dacewar fasahar don takamaiman aikace-aikace.
5. Aikace-aikace na Gaba & Hanyoyin Bincike
EMGesture tana buɗe hanyar aikace-aikace masu ƙirƙira da yawa:
- Mota: Sarrafa motsin hannu don tsarin nishadi daga kushin caji mara waya na tsakiyar kwamfutar, yana rage shagaltuwar direba.
- Gidaje masu Wayo: Sarrafa fitilu, kiɗa, ko na'urori ta hanyar motsin hannu akan cajin gefen gado ko tebur.
- Samun dama: Bayar da mu'amalar sarrafawa ba tare da taɓawa ba ga mutanen da ke da nakasu na motsi.
- Kiosk na Jama'a/Kanti: Hulɗa mai tsafta, ba tare da taɓawa ba tare da nuni na bayanai ko tashoshin biyan kuɗi.
Hanyoyin Bincike na Gaba:
- Ƙaddamar da Kewayo & Ganowa na 3D: Yin amfani da coils na caji da yawa ko tsararrun lokaci don ƙaddamar da kewayon ganowa da ba da damar bin diddigin motsin hannu na 3D.
- Keɓance Motsin Hannu & Daidaitawa: Aiwatar da koyo akan na'ura don ba da damar masu amfani su ayyana al'ada na motsin hannu da daidaitawa da salon mutum ɗaya.
- Haɗa Nau'i-nau'i: Haɗa bayanan motsin hannu na EM tare da mahallin daga wasu na'urori masu gano (misali, na'urar auna sauri na na'ura, hasken muhalli) don warware niyyoyi da ba da damar ƙarin rikitarwar hulɗa.
- Daidaituwa & Tsaro: Haɓaka ƙa'idodi don tabbatar da tsaron bayanan motsin hannu da hana mugun yaudarar siginonin EM.
6. Nassoshi
- Wang, W., Yang, L., Gan, L., & Xue, G. (2025). Cajin Mara Waya a matsayin Na'urar Gano Motsin Hannu: Sabuwar Hanya zuwa Hulɗa Ko'ina. A cikin Proceedings of CHI Conference on Human Factors in Computing Systems (CHI '26).
- Hukumar Kula da Tsaron Tituna ta Amurka (NHTSA). (2023). Bayanan Mutuwar Direba da ke da Hankali.
- Zhu, H., et al. (2020). Damuwar Sirri a cikin Gane Ayyukan Mutum na Tushen Kyamara: Bincike. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies.
- Grand View Research. (2023). Rahoton Girman Kasuwar Hulɗar Mutum da Injin.
- Zhang, N., et al. (2021). Mataimakin Muryar ku Nawa ne: Yadda ake Yin Amfani da Lasifikai don Sace Bayanai da Sarrafa Wayar ku. A cikin Proceedings of the ACM SIGSAC Conference on Computer and Communications Security.
- Yang, L., et al. (2023) Ganowa na Mutum na Tushen RF: Daga Gane Motsin Hannu zuwa Saka idanu Alamar Rayuwa. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies.
- Zhu, J.-Y., Park, T., Isola, P., & Efros, A. A. (2017). Fassarar Hoto zuwa Hoto mara Biyu ta amfani da Cibiyoyin Adawa na Da'ira-Daidaitacce. A cikin Proceedings of the IEEE International Conference on Computer Vision (ICCV).
- IEEE Xplore Digital Library. Takardun asali akan Ganowa da Ƙirar Lantarki.
- Cibiyar Ƙididdiga da Fasaha ta Ƙasa (NIST). Rahotanni kan Kimanta Tsarin Ganowa.