sabman/aicasts.com

Weekly AI Tutorials

Hi there 👋 🧐

📫 How to reach me:

• https://twitter.com/sabman
• https://burq.fun

Learnings

• git
• golang
• Engineering Management
• Entrepreneurship
• Serving on (iNGO) Boards
• Mental Health
• K8s
• CS Intern Guide

aicasts.com's Issues

geotrellis-landsat-tutorial

https://github.com/geotrellis/geotrellis-landsat-tutorial

dagster

https://github.com/dagster-io/dagster

flatgeobuf

https://github.com/bjornharrtell/flatgeobuf

cuspatial

https://github.com/rapidsai/cuspatial

[big-data] test presto's geospatial capabilities

see: prestodb/presto#12478

try ludwig from uber

https://ludwig-ai.github.io/ludwig-docs
https://github.com/uber/ludwig

Ludwig is a toolbox built on top of TensorFlow that allows to train and test deep learning models without the need to write code.

[modelling] AgentMaps

https://github.com/noncomputable/AgentMaps

Streamlit

https://docs.streamlit.io/en/stable/

[BigData][GPU][database] try omniscidb

https://github.com/omnisci/omniscidb
https://github.com/omnisci/omniscidb/wiki/Tutorials-&-Demos

[data-privacy][pg] postgresql-anonymizer

https://postgresql-anonymizer.readthedocs.io/en/latest/

[pipelines] metaflow

https://github.com/Netflix/metaflow

Geomesa on Kafka

Official: https://www.geomesa.org/

• Talks about Geomesa
• Client https://www.pubnub.com/blog/realtime-geo-tracking-app-react-native/
• React Native https://github.com/transistorsoft/react-native-background-geolocation

try Apache Beam

https://www.youtube.com/watch?v=wlrvooFLlXo

[spatial stats] try pysal

VROOM-Project

https://github.com/VROOM-Project/vroom
https://www.youtube.com/watch?time_continue=4&v=5i7xs7FiIvs

[OpenData] Geospatial

• https://grid3.org/resources/data
• https://www.maxar.com/open-data
• http://www.data4sdgs.org/resource-listing
• https://portal.opentopography.org/datasets

[DataCollection] Tools

• http://openmapkit.org/
• http://opendatakit.org/
• https://qfield.org/

herepy

https://herepy.abdullahselek.com/

Available clients are:

• Destination Weather API
• Geocoder API
• Geocoder Autocomplete API
• Reverse Geocoder API
• Places API
• Public Transit API
• Fleet Telematics Route Matching API
• Routing API
• EV Charging Stations API
• Fleet Telematics API

[GeoAI][DeepLearning] TorchSat

https://torchsat.readthedocs.io/en/latest/

Deep Learning in Medical Imaging

• https://github.com/deeponcology/applied-dl-2018
• https://towardsdatascience.com/deep-learning-and-medical-image-analysis-for-malaria-detection-with-fastai-c8f08560262f
• https://www.kaggle.com/saburq/hello-world-for-deep-learning-siim/
• Hello World Deep Learning in MI https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5959832
• Kaggle Notebook for the paper Hello World Deep Learning in MI https://www.kaggle.com/wfwiggins203/hello-world-for-deep-learning-siim

[AR] AR-js

https://github.com/AR-js-org/AR.js

[usecases] Bushfire modelling software: CSIRO spark

• https://research.csiro.au/static/spark/Spark_applications_user_guide_v112.pdf
• https://research.csiro.au/spark/download/

[API] Try Insomnia Designer

https://konghq.com/blog/introducing-insomnia-designer-an-open-source-collaborative-editor-for-api-design/?itm_source=website&itm_medium=news-promo

[ml-platform] https://flyte.org/

• https://github.com/lyft/flyte

from release blog post

The problems Flyte solves

With data now being a primary asset for companies, executing large-scale compute jobs is critical to the business, but problematic from an operational standpoint. Scaling, monitoring, and managing compute clusters becomes a burden on each product team, slowing down iteration and subsequently product innovation. Moreover, these workflows often have complex data dependencies. Without platform abstraction, dependency management becomes untenable and makes collaboration and reuse across teams impossible.

Flyte’s mission is to increase development velocity for machine learning and data processing by abstracting this overhead. We make reliable, scalable, orchestrated compute a solved problem, allowing teams to focus on business logic rather than machines. Furthermore, we enable sharing and reuse across tenants so a problem only needs to be solved once. This is increasingly important as the lines between data and machine learning converge, including the roles of those who work on them.

automating-gis-processes Course

https://automating-gis-processes.github.io/site/

[big geodata] vectorpipe

https://github.com/geotrellis/vectorpipe

[postgresql-extension] google's differential_privacy

https://github.com/google/differential-privacy/tree/master/differential_privacy

[Geo Data Spec] Cloud Optimized GeoTiff

investigate the use of https://github.com/cogeotiff for big data satellite image processing

• Implementations: https://www.cogeo.org/#implementations
• Spark Data Frames: https://rasterframes.io/
• Javascript Client: https://geotiffjs.github.io/

[geo] explore movingpandas

http://exploration.movingpandas.org/

Workflow Automation https://n8n.io/

[mllib] Transformers for NLP

Docs

https://huggingface.co/transformers/

Why?

Philosophy

[Simulation] Mobility Data ABStreet

https://github.com/dabreegster/abstreet/#ab-street

[stack] postgraphile starter

https://github.com/graphile/starter

[visualization] idyll-lang

https://idyll-lang.org/

compare mlops tooling

1. Uber’s Michelangelo
2. Facebook's FBLearner Flow
3. Google's TFX
4. Databricks' MLFlow

[task] Understand and run OpenDroneMap as a service

https://github.com/OpenDroneMap/WebODM#getting-started

• run the server
• collect data
• process data on WebODM

[inference] Try TensorRT

https://developer.nvidia.com/tensorrt
Example: https://devblogs.nvidia.com/speed-up-inference-tensorrt/

geotrellis

https://geotrellis.io/

• raster
• scala/akka

try rocm/hopsworks

https://rocm.github.io/
https://hops.readthedocs.io/en/0.9/index.html
https://www.youtube.com/watch?v=neb1C6JlEXc

ASPRS LAS Specification

https://github.com/ASPRSorg/LAS
ASPRS LAS Specification, the remote sensing industry's most widely used file format for point cloud data

try AutoML

• https://www.automl.org/

[3D][Geo] 3d-tiles Spec

3D tiles spec is an emerging spec to bring fast 3D data geospatial visualization to the web. Similar to the way Victor Tiles did for non-3D data. https://github.com/CesiumGS/3d-tiles

ibi-transit data tools

https://data-tools-docs.ibi-transit.com/en/latest/dev/deployment/

Geospatial Anomaly Detection

Geospatial Anomaly Detection
https://www.instaclustr.com/geospatial-anomaly-detection-with-kafka-cassandra/

try taxi dataset

https://github.com/toddwschneider/nyc-taxi-data
https://toddwschneider.com/posts/analyzing-1-1-billion-nyc-taxi-and-uber-trips-with-a-vengeance/

[ML platform] Try BentoML

https://docs.bentoml.org/en/latest/

try apache flink

https://github.com/ververica/flink-training-exercises
https://github.com/ververica/sql-training
https://bdaschool2019.github.io/files/flink.pdf
https://training.ververica.com/setup/taxiData.html

[postgres-extension] mobilitydb

• https://docs.mobilitydb.com/nightly/
• https://docs.mobilitydb.com/nightly/workshop/
• https://github.com/ULB-CoDE-WIT/MobilityDB

[visualization] try martin vector tile server

https://github.com/urbica/martin

test map matching with barefoot

Barefoot is a java application with ML based map matching algorithms
https://github.com/bmwcarit/barefoot/wiki#tracker-server
API
https://github.com/bmwcarit/barefoot
(overlaps features with osrm)