Deep Learning Specialization

Master CNNs, RNNs, Transformers, and generative models — with projects you can ship and showcase.

Deep Learning · Intermediate · 12 weeks · 36 lessons · Free

Prereq: comfortable with basic neural nets and Python.

Modules:

• CNNs: modern architectures (ResNet/EfficientNet), augmentation, transfer learning.

• RNNs: LSTM/GRU for sequences and time‑series.

• Transformers: attention, encoder/decoder; apply pre‑trained models.

• Generative models: VAEs, GANs, diffusion.

Projects: image classifier, detector, segmenter; text tasks; generative demos. Portfolio: 8 solid projects.

Lessons

  1. CNNs: modern architectures (ResNet, EfficientNet) and transfer learning (~55 min)
  2. Data augmentation and regularization (~35 min)
  3. RNNs and LSTMs for sequences and time-series (~50 min)
  4. Transformers: attention and encoder/decoder (~60 min)
  5. Generative models: VAEs, GANs, and diffusion (~55 min)
  6. Portfolio projects: classifier, detector, segmenter (~45 min)