BiteSizedChunks.comLearn one small thing at a time.

Course contentsShow

Machine Learning and Deep Learning

931The Image Classification Problem
932ImageNet and the Data Revolution
933Why Pretrained Models Work
934Feature Hierarchy in CNNs
935Transfer Learning Fundamentals
936Fine-Tuning vs Feature Extraction
937Layer Freezing Strategies
938Learning Rate Considerations for Fine-Tuning
939Data Augmentation for Classification
940Handling Small Datasets with Transfer Learning
941Domain Adaptation Challenges
942Multi-Task and Multi-Domain Learning
943Model Selection for Transfer Learning
944Implementing Transfer Learning in PyTorch
945Object Detection vs Classification
946Bounding Box Representations
947Intersection over Union (IoU)
948Object Detection as Classification + Localization
949Anchor Boxes Concept
950The Sliding Window Approach
951Region Proposal Methods
952Two-Stage vs One-Stage Detectors
953Multi-Scale Feature Detection
954Non-Maximum Suppression (NMS)
955R-CNN Architecture
956Fast R-CNN Improvements
957Region of Interest (RoI) Pooling
958Detection Loss Functions
959Class Imbalance in Object Detection
960Mean Average Precision (mAP)
961From Two-Stage to One-Stage: The YOLO Revolution
962YOLO Architecture: Grid-Based Detection
963YOLO Loss Function: Balancing Multiple Objectives
964YOLOv2 and YOLOv3: Incremental Improvements
965YOLOv4 and YOLOv5: Speed and Accuracy Advances
966YOLOX: Anchor-Free and Decoupled Head
967YOLOv7 and YOLOv8: State-of-the-Art Real-Time Detection
968SSD: Multi-Scale Feature Maps for Detection
969RetinaNet and Focal Loss
970EfficientDet: Compound Scaling for Detection
971DETR: Detection with Transformers
972Deformable DETR: Efficient Attention for Detection
973Modern Detection Trade-offs: Speed vs Accuracy
974Post-Processing: NMS Variants and Soft-NMS
975What Is Semantic Segmentation
976Segmentation Output Format and Evaluation
977Fully Convolutional Networks (FCN)
978Upsampling and Transposed Convolutions
979U-Net Architecture
980Skip Connections in Segmentation Networks
981DeepLab and Atrous Convolutions
982Atrous Spatial Pyramid Pooling (ASPP)
983Loss Functions for Segmentation
984Semantic Segmentation Datasets
985Multi-Scale Inference and Test-Time Augmentation
986Segmentation Model Design Trade-offs
987Instance Segmentation Overview
988Mask R-CNN Architecture
989Mask R-CNN Training and Loss Functions
990ROI Align vs ROI Pooling
991Panoptic Segmentation
992Keypoint Detection and Pose Estimation
993Image Captioning Fundamentals
994Visual Question Answering (VQA)
995Video Understanding Tasks
996Optical Flow and Motion Estimation
997Depth Estimation from Single Images
9983D Object Detection and Point Clouds

Machine Learning and Deep Learning

931The Image Classification Problem
932ImageNet and the Data Revolution
933Why Pretrained Models Work
934Feature Hierarchy in CNNs
935Transfer Learning Fundamentals
936Fine-Tuning vs Feature Extraction
937Layer Freezing Strategies
938Learning Rate Considerations for Fine-Tuning
939Data Augmentation for Classification
940Handling Small Datasets with Transfer Learning
941Domain Adaptation Challenges
942Multi-Task and Multi-Domain Learning
943Model Selection for Transfer Learning
944Implementing Transfer Learning in PyTorch
945Object Detection vs Classification
946Bounding Box Representations
947Intersection over Union (IoU)
948Object Detection as Classification + Localization
949Anchor Boxes Concept
950The Sliding Window Approach
951Region Proposal Methods
952Two-Stage vs One-Stage Detectors
953Multi-Scale Feature Detection
954Non-Maximum Suppression (NMS)
955R-CNN Architecture
956Fast R-CNN Improvements
957Region of Interest (RoI) Pooling
958Detection Loss Functions
959Class Imbalance in Object Detection
960Mean Average Precision (mAP)
961From Two-Stage to One-Stage: The YOLO Revolution
962YOLO Architecture: Grid-Based Detection
963YOLO Loss Function: Balancing Multiple Objectives
964YOLOv2 and YOLOv3: Incremental Improvements
965YOLOv4 and YOLOv5: Speed and Accuracy Advances
966YOLOX: Anchor-Free and Decoupled Head
967YOLOv7 and YOLOv8: State-of-the-Art Real-Time Detection
968SSD: Multi-Scale Feature Maps for Detection
969RetinaNet and Focal Loss
970EfficientDet: Compound Scaling for Detection
971DETR: Detection with Transformers
972Deformable DETR: Efficient Attention for Detection
973Modern Detection Trade-offs: Speed vs Accuracy
974Post-Processing: NMS Variants and Soft-NMS
975What Is Semantic Segmentation
976Segmentation Output Format and Evaluation
977Fully Convolutional Networks (FCN)
978Upsampling and Transposed Convolutions
979U-Net Architecture
980Skip Connections in Segmentation Networks
981DeepLab and Atrous Convolutions
982Atrous Spatial Pyramid Pooling (ASPP)
983Loss Functions for Segmentation
984Semantic Segmentation Datasets
985Multi-Scale Inference and Test-Time Augmentation
986Segmentation Model Design Trade-offs
987Instance Segmentation Overview
988Mask R-CNN Architecture
989Mask R-CNN Training and Loss Functions
990ROI Align vs ROI Pooling
991Panoptic Segmentation
992Keypoint Detection and Pose Estimation
993Image Captioning Fundamentals
994Visual Question Answering (VQA)
995Video Understanding Tasks
996Optical Flow and Motion Estimation
997Depth Estimation from Single Images
9983D Object Detection and Point Clouds

← Machine Learning and Deep Learning

Lesson 993 of 3,538·22. Computer Vision TasksPro lesson

Image Captioning Fundamentals

Generating textual descriptions from images: combining CNN encoders with language model decoders for automatic caption generation.

This lesson is for subscribers

You've completed the free preview. Subscribe to unlock every lesson in every course.

See pricing Back to course