Ikibazo gihuza ibibazo · Hagati Urwego

Gutandukanya Imvugo hamwe n'ikibazo cya Cocktail Ikibazo

Reba uburyo sisitemu yo gutandukanya imvugo ikoresha amajwi avanze, ibimenyetso byerekana umwanya, kubara abavuga, hamwe no gukuramo intego-mu majwi afatika.

Ikibazo 1 Bya 8

Ni ikihe 'kibazo cy'ishyaka rya cocktail'?

Reba buri kibazo8 ibibazo

Ibisubizo nibisobanuro bigaragara mugihe ufata ikibazo.

  1. Ni ikihe 'kibazo cy'ishyaka rya cocktail'?
    • Kwibanda kumajwi imwe cyangwa isoko yijwi hagati yandi menshi arushanwa
    • Guhitamo ibinyobwa bikwiye mubirori
    • Gupima amajwi y'icyumba
    • Guhagarika umuziki w'ishyaka kugirango ugende
  2. Nibiki bisohoka muri sisitemu yo gutandukanya imvugo ihabwa amajwi avanze y'abavuga benshi?
    • Indangururamajwi imwe yo kuvanga
    • Ijwi ryihariye, ryitaruye amajwi kuri buri muvugizi
    • Incamake yinyandiko gusa
    • MP3 ifunitse yicyumba cyose
  3. Niki Conv-TasNet ikora muburyo butandukanye nuburyo bwinshi bwo gutandukana?
    • Irasaba byibuze mikoro umunani
    • Ikora gusa kumuziki wanditswe mbere
    • Ikora mu buryo butaziguye mugihe cyateganijwe kuri raw waveform sample aho kuba spekrogramu
    • Ihindura imvugo mbere yo kuyitandukanya
  4. Nigute imiyoboro myinshi yo gutandukanya itandukanya ijwi ryumuntu kuvanga?
    • Mu kongera igipimo cyicyitegererezo
    • Mugusiba imirongo yose iri hejuru ya 1 kHz
    • Muguhindura amajwi kumurongo
    • Mugereranije mask itoranya buri muvugizi igice cyikimenyetso
  5. 'Gukuramo intego-kuvuga' ni iki?
    • Gukuramo umuntu umwe wihariye ukoresheje urugero rugufi rw'ijwi ryabo
    • Kuraho amajwi yose yabantu mumajwi
    • Kubara abavuga bangahe
    • Kuzamura umuziki winyuma
  6. Kuki kongera amashusho bifasha gutandukanya imvugo?
    • Video igabanya ingano ya dosiye y amajwi
    • Kugenda kwiminwa bitanga ibimenyetso bifasha guhuza amajwi kubavuga neza
    • Kamera zihagarika echo mu buryo bwikora
    • Video ikuraho ibikenewe byose gutunganya amajwi
  7. What makes overlapping speech difficult for a separation system before it has analyzed the mixture?
    • The mixture contains no audio samples
    • Every speaker always occupies a separate frequency band
    • The number of speakers and which parts of the waveform belong to each speaker may be unknown
    • The recording already labels each word with its speaker
  8. What cue can an early multi-microphone separation method use to distinguish speakers?
    • The spelling of each speaker's name
    • The color of the recording interface
    • The file's compression ratio alone
    • Spatial differences in how each voice reaches the microphone array