Update README.md

Author: robmarkcole

README.md

@@ -330,6 +330,7 @@ Several different techniques can be used to count the number of objects in an im
 * [Detecting-Cyclone-Centers-Custom-YOLOv3](https://github.com/ShubhayanS/Detecting-Cyclone-Centers-Custom-YOLOv3) -> tropical cyclones (TCs) are intense warm-corded cyclonic vortices, developed from low-pressure systems over the tropical oceans and driven by complex air-sea interaction
 * [Object-Detection-YoloV3-RetinaNet-FasterRCNN](https://github.com/bostankhan6/Object-Detection-YoloV3-RetinaNet-FasterRCNN) -> trained on a private datset
 * [Google-earth-Object-Recognition](https://github.com/InnovAIco/Google-earth-Object-Recognition) -> Code for training and evaluating on Dior Dataset (Google Earth Images) using RetinaNet and YOLOV5
+* [AI-TOD](https://github.com/jwwangchn/AI-TOD) -> a dataset for tiny object detection in aerial images. The mean size of objects in AI-TOD is about 12.8 pixels, which is much smaller than other datasets
 
 #### Object detection enhanced by super resolution
 * [Super-Resolution and Object Detection](https://medium.com/the-downlinq/super-resolution-and-object-detection-a-love-story-part-4-8ad971eef81e) -> Super-resolution is a relatively inexpensive enhancement that can improve object detection performance
@@ -359,6 +360,7 @@ Several different techniques can be used to count the number of objects in an im
 * [OHDet_Tensorflow](https://github.com/SJTU-Thinklab-Det/OHDet_Tensorflow) -> can be applied to rotation detection and object heading detection
 * [Seodore](https://github.com/nijkah/Seodore) -> framework maintaining recent updates of mmdetection
 * [Rotation-RetinaNet-PyTorch](https://github.com/HsLOL/Rotation-RetinaNet-PyTorch) -> oriented detector Rotation-RetinaNet implementation on Optical and SAR ship dataset
+* [AIDet](https://github.com/jwwangchn/aidet) -> an open source object detection in aerial image toolbox based on MMDetection
 
 #### Object detection - buildings, rooftops & solar panels
 * [Machine Learning For Rooftop Detection and Solar Panel Installment](https://omdena.com/blog/machine-learning-rooftops/) discusses tiling large images and generating annotations from OSM data. Features of the roofs were calculated using a combination of contour detection and classification. [Follow up article using semantic segmentation](https://omdena.com/blog/rooftops-classification/)
@@ -476,6 +478,7 @@ Generally treated as a semantic segmentation problem or custom features created
 * [fcd](https://github.com/jnyborg/fcd) -> code for 2021 paper: Fixed-Point GAN for Cloud Detection. A weakly-supervised approach, training with only image-level labels
 * [CloudX-Net](https://github.com/sumitkanu/CloudX-Net) -> an efficient and robust architecture used for detection of clouds from satellite images
 * [A simple cloud-detection walk-through using Convolutional Neural Network (CNN and U-Net) and fast.ai library](https://medium.com/analytics-vidhya/a-simple-cloud-detection-walk-through-using-convolutional-neural-network-cnn-and-u-net-and-bc745dda4b04)
+* [38Cloud-Medium](https://github.com/cordmaur/38Cloud-Medium) -> Walk-through using u-net to detect clouds in satellite images with fast.ai
 
 ## Change detection & time-series
 Monitor water levels, coast lines, size of urban areas, wildfire damage, crop changes. Note, clouds change often too..!
@@ -1070,6 +1073,10 @@ This section contains a short list of datasets relevant to deep learning, partic
 * https://downloads.greyc.fr/vedai/
 * [pytorch-vedai](https://github.com/MichelHalmes/pytorch-vedai) -> object detection on the VEDAI dataset: Vehicle Detection in Aerial Imagery
 
+## AI-TOD - tiny object detection
+* https://github.com/jwwangchn/AI-TOD
+* The mean size of objects in AI-TOD is about 12.8 pixels, which is much smaller than other datasets
+
 ## AIRS (Aerial Imagery for Roof Segmentation)
 * https://www.airs-dataset.com
 * Public dataset for roof segmentation from very-high-resolution aerial imagery (7.5cm)
@@ -1090,6 +1097,10 @@ This section contains a short list of datasets relevant to deep learning, partic
 * Imagery appears to be global but with significant fraction from North America
 * Winning solution published by neptune.ai [here](https://github.com/neptune-ai/open-solution-mapping-challenge), achieved precision 0.943 and recall 0.954 using Unet with Resnet.
 
+## BONAI - building footprint dataset
+* https://github.com/jwwangchn/BONAI
+* BONAI (Buildings in Off-Nadir Aerial Images) is a dataset for building footprint extraction (BFE) in off-nadir aerial images
+
 ## GID15 large scale semantic segmentation dataset
 * https://captain-whu.github.io/GID15/