Deep Learning, Computer Vision, and Natural Language Processing!
Curriculum
Available in
days
days
after you enroll
Available in
days
days
after you enroll
Available in
days
days
after you enroll
- 1 - Calculus (25:07)
- 2 - Basics of Gradient Descent (62:56)
- 3 - Coding Gradient Descent and Understanding Learning Rates (33:46)
- 4 - Versions of GD, Minimas, Initialisation, Feature Scaling, Learning Rate (43:56)
- 5 - Why Tensorflow? (31:34)
- 6 - Project: Intro to Object Segmentation (52:52)
- 7 - Project: Preprocessing (30:56)
- 8 - Project: Model Building and Evaluation (63:23)
- Refresher Quiz (Non-graded)
Available in
days
days
after you enroll
Available in
days
days
after you enroll
- Day 1 (a) - Intro to DL, Perceptron, MLP (34:50)
- Day 1 (b) - Deep Learning v/s Classical ML (28:44)
- Day 2 - How NN learn, Notation, Forward Pass (45:51)
- Day 3 - Trainable Parameters, Backpropagation Derivation, and Memoization (40:36)
- Day 4 - Let's code it up! (40:49)
- Day 5 - What are the correct number of epochs? (30:10)
- Day 6 & 7 - Quiz 2
- Day 8 - Intro to Activation Functions & The Important of Non-linearity (41:02)
- Day 9 (a) - Different Activation Functions with properties (33:58)
- Day 9 (b) - Dying ReLU: what, why, and the solution! (23:33)
- Day 9 (c) - Variants of ReLU, SoftMax (16:17)
- Day 10 - Loss Functions (56:58)
- Day 11 - Recap of Information Theory (79:00)
Available in
days
days
after you enroll
- Day 1 - What is the correct batch size? (43:22)
- Day 2 - Vanishing Gradient: Problem (35:31)
- Day 3 - (a) Vanishing Gradient: Detection & Solution (14:47)
- Day 3 - (b) Exploding Gradient Problem & Gradient Clipping (20:18)
- Day 4 - (a) Weight Initialisation (22:31)
- Day 4 - (b) Code: Activation Fns, Vanishing/Exploding Grad, Weight Init (29:29)
- Day 5 - Depth v/s Width (18:46)
Available in
days
days
after you enroll
- Day 1 - Color Spaces and Types of Images (47:41)
- Day 2 - Image Manipulation with Filters (53:55)
- Day 3 - OpenCV: Color Spaces, Equivalisation, and Thresholding (45:12)
- Day 4 - OpenCV: Image Smoothing, Morphological Ops, Edge Detection (26:46)
- Day 5 - Quiz 3
- Day 6 - How CNNs were built? (31:20)
- Day 7 (a) - Convolution, Feature Maps, and Output Sizes (35:54)
- Day 7 (b) - Parameter Sharing, Num of Params, Receptive Field (23:58)
- Day 8 (a) - Visualising the feature maps (26:30)
- Day 8 (b) - Let's play pool! (32:45)
- Day 9 - Coding a CNN, CNN v/s ANN on Cifar-10, Softmax Stability, Evolution of CNNs (LeNet, AlexNet, etc.) (61:54)
- Day 10 & 11 - Model Building: Pre-existing, Pre-training, Transfer, Finetune, Auxilliary, Unsupervised Pre-training (75:39)
- Day 12 & 13 - Quiz 4: Build a CNN on the Cifar-100 dataset
- Day 14 - Data Augmentation (40:24)
- Day 15 - Overfitting and Dropout (62:14)
- Day 16 (a) - Implementing Dropout (8:29)
- Day 16 (b) - Regularisation in Neural Networks (12:07)
- Day 17 - ResNet, Residual Blocks, and Skip Connections (34:26)
- Day 18 - Encoder-Decoder, Convolution Transpose, U-net (42:49)
Available in
days
days
after you enroll
Available in
days
days
after you enroll
Available in
days
days
after you enroll
Available in
days
days
after you enroll
Available in
days
days
after you enroll
Available in
days
days
after you enroll
- Seminal Papers List
- 1.1: Introduction to Research in AI (111:53)
- 1.2: Submit Your Literature Survey!
- 2.1: 9 June - Live Session on Branding Beyond Github Profiles! (28:11)
- 2.2: Submit Your Medium Blog!
- 3.1 - Lab Session: Let's Build A Virtual TryOn! (Model Designs, Inpainting, Running GitHub Repos, Custom Data Loaders & Losses) (207:54)