일 | 월 | 화 | 수 | 목 | 금 | 토 |
---|---|---|---|---|---|---|
1 | ||||||
2 | 3 | 4 | 5 | 6 | 7 | 8 |
9 | 10 | 11 | 12 | 13 | 14 | 15 |
16 | 17 | 18 | 19 | 20 | 21 | 22 |
23 | 24 | 25 | 26 | 27 | 28 |
Tags
- 모션매칭
- 언리얼엔진
- motion matching
- multimodal
- userwidget
- 생성모델
- deep learning
- Stat110
- Diffusion
- WinAPI
- RNN
- animation retargeting
- cv
- 오블완
- BERT
- Unreal Engine
- GAN
- dl
- WBP
- 딥러닝
- 디퓨전모델
- CNN
- NLP
- Generative Model
- 폰트생성
- ue5.4
- Few-shot generation
- ddpm
- UE5
- Font Generation
Archives
- Today
- Total
목록dcgan (1)
Deeper Learning
Deep Convolutional Generative Adversarial Networks (DCGAN)
DCGAN GAN(Generative Adversarial Networks)은 Generator와 Discriminator 2개의 모델을 사용하여 실제 데이터와 비슷한 데이터를 생성하는 생성 모델이다. Random 한 Noise vector를 input으로 사용하는 Generator는 실제 데이터와 같은 shape의 데이터를 output으로 한다. Discriminator는 실제 데이터와, Generator가 생성한 데이터를 구별해내는 모델로 Binary Classification 모델이다. DCGAN은 Dense Net이 아닌 Convolutional NN를 사용하는 모델이다. DCGAN structure Generator는 위와 같은 구조로 위의 예시는 100d의 noise vector는 Transp..
AI/Deep Learning
2020. 12. 28. 21:47