MetNet Neural Weather Model: Difference between revisions

Latest revision as of 16:39, 31 January 2022

MetNet: A Neural Weather Model for Precipitation Forecasting
Authors: Casper Kaae Sønderby, Lasse Espeholt, Jonathan Heek, Mostafa Dehghani, Avital Oliver, Tim Salimans, Shreya Agrawal, Jason Hickey, Nal Kalchbrenner

Paper
Google AI Blog

Model Architecture

The MetNet model consists of 3 parts

Spatial Downsampler
Temporal Encoder (Conv LSTM ^[1])
Spatial Aggregator (Axial Attention ^[2])

Resources

Yannic Kilcher Video on MetNet and Axial Attention

References

↑ Xingjian Shi, Zhourong Chen, Hao Wang, Dit-Yan Yeung, Wai-kin Wong, Wang-chun Woo (2015). Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting
↑ Jonathan Ho, Nal Kalchbrenner, Dirk Weissenborn, Tim Salimans (2019). Axial Attention in Multidimensional Transformers

[xingjian2015convlstm-1] Xingjian Shi, Zhourong Chen, Hao Wang, Dit-Yan Yeung, Wai-kin Wong, Wang-chun Woo (2015). Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting

[ho2019axial-2] Jonathan Ho, Nal Kalchbrenner, Dirk Weissenborn, Tim Salimans (2019). Axial Attention in Multidimensional Transformers

[1]

[2]

Revision as of 14:31, 8 April 2020 view source David (talk \| contribs) Bureaucrats, Interface administrators, Administrators 5,418 edits Created page with "MetNet: A Neural Weather Model for Precipitation Forecasting<br> Authors: Casper Kaae Sønderby, Lasse Espeholt, Jonathan Heek, Mostafa Dehghani, Avital Oliver, Tim Salimans,..."		Latest revision as of 16:39, 31 January 2022 view source David (talk \| contribs) Bureaucrats, Interface administrators, Administrators 5,418 edits No edit summary
Line 18:		Line 18:

	==References==		==References==
			[[Category: Papers]]