/DLCV

CommE5052 Deep Learning for Computer Vision

Primary LanguagePythonMIT LicenseMIT

CommE5052 Deep Learning for Computer Vision

HW1

  • Image classification
  • Semantic segmentation

HW2

  • Image Generation: GAN/Diffusion models
  • Unsupervised Domain Adaptation: DANN

HW3

  • Zero-shot image Classification with CLIP
  • Image Captioning with VL-model
  • Visualization of Attention in Image Captioning

HW4

  • 3D Novel View Synthesis
  • Self-supervised Pre-training for Image Classification