Kuis terkait panduan · Sedang Tingkat

Kuis Gambar Teks-ke-Gambar

Tinjau pengondisian teks Imagen asli, pembuatan berjenjang, ambang batas dinamis, dan pilihan ketahanan dalam pelatihan resolusi super.

Jalur panduan terkaitTeks Gambar Ke Gambar
Pertanyaan 1 dari 8

In the original Imagen study's scaling comparison, which change produced larger image-quality and text-alignment gains?

Pratinjau setiap pertanyaan8 pertanyaan

Jawaban dan penjelasan muncul saat Anda mengikuti kuis.

  1. In the original Imagen study's scaling comparison, which change produced larger image-quality and text-alignment gains?
    • Discarding the language encoder entirely
    • Increasing the pretrained text encoder's size rather than the diffusion model's size
    • Replacing diffusion with nearest-neighbor image lookup
    • Reducing every prompt to a single word
  2. Encoder teks mana yang diandalkan Imagen?
    • Jaringan gambar konvolusional
    • LSTM kecil
    • Seperangkat aturan yang ditulis tangan
    • Model bahasa T5-XXL yang dibekukan
  3. Bagaimana Imagen mencapai resolusi tinggi?
    • Difusi bertingkat dengan tahapan peningkatan resolusi super
    • Sebuah langkah generasi raksasa
    • Dengan memotong foto stok berukuran besar
    • Dengan menginterpolasi dua gambar
  4. What problem did original Imagen's dynamic thresholding address under strong guidance?
    • A lack of text tokens in the encoder vocabulary
    • The need to compress the final PNG file
    • Saturation of predicted pixel values during sampling
    • A missing training label for every object
  5. Apa itu DrawBench?
    • Tablet gambar
    • Merek GPU
    • Kumpulan data foto kucing
    • Tolok ukur petunjuk yang menantang untuk mengevaluasi model teks-ke-gambar
  6. Mengapa encoder T5-XXL tetap dibekukan di Imagen?
    • Itu rusak dan tidak bisa dilatih
    • Pemahaman teksnya yang kaya dari transfer pra-pelatihan tanpa penyetelan lebih lanjut
    • Untuk menghemat ruang disk saja
    • Karena teks tidak relevan
  7. What resolution did original Imagen's base diffusion model generate before its super-resolution stages?
    • 256 by 256 pixels as the first stage
    • 1024 by 1024 pixels directly
    • 64 by 64 pixels
    • A single pixel that was copied across the final image
  8. Why did the Imagen researchers add Gaussian noise to low-resolution conditioning images during super-resolution training?
    • To make the upsamplers more robust to imperfect inputs from earlier stages
    • To remove text conditioning from the entire cascade
    • To guarantee that all training images had the same caption
    • To replace the upsamplers with deterministic resizing