Tensorflow-io

Major Features
* Added support for stream timeout in KafkaGroupIODataset
* Renamed gstpufs to gsmemcachedfs
* Added experimental KakfaBatchIODataset for online learning
* Added gabor/laplacian/gaussian filter support
* Added basic Arrow string type support for BigQuery reader
* Added ElasticSearch dataset support
* Added Prewitt and Sobel filter support
* Added arbitary dimensional support for `decode_json`
* Added bool data type support for `decode_json`
* Added tutorials for Kafka and ElasticSearch usage with tf.keras.
* Containers of tensorflow-io releases are available
at `docker pull tfsigio/tfio`.

Thanks to our Contributors

This release contains contributions from many people:

Aleksey Vlasenko, Daniel, Kota Yamaguchi, Mark Daoust,
Michael Kuchnik, Vignesh Kothapalli, Yong Tang, Yuan Tang,
emkornfield, marioecd

We are also grateful to all who filed issues or helped resolve them, asked and
answered questions, and were part of inspiring discussions.

Major Features
* Add basic gstpu file system for GCS/TPU support.
* Add null handling support for `AvroRecordDataset`.
* Add tfio.experimental.color for colorspace conversions:
- `tfio.experimental.color.[rgb_to_bgr|bgr_to_rgb]`: BGR
- `tfio.experimental.color.[rgb_to_rgba|rgba_to_rgb]`: RGBA
- `tfio.experimental.color.[rgb_to_ycbcr|ycbcr_to_rgb]`: YCbCr
- `tfio.experimental.color.[rgb_to_ypbpr|ypbpr_to_rgb]`: YPbPr
- `tfio.experimental.color.[rgb_to_ydbdr|ydbdr_to_rgb]`: YDbDr
- `tfio.experimental.color.[rgb_to_xyz|xyz_to_rgb]`: CIE XYZ
- `tfio.experimental.color.[rgb_to_lab|lab_to_rgb]`: CIE LAB
- `tfio.experimental.color.[rgb_to_yiq|yiq_to_rgb]`: YIQ
- `tfio.experimental.color.[rgb_to_yuv|yuv_to_rgb]`: YUV
- `tfio.experimental.color.rgb_to_grayscale`: Grayscale(BT 709)
* Add `tfio.image.encode_gif` for GIF (animated) encoding support.

Thanks to our Contributors

This release contains contributions from many people:

Cheng Ren, Paul Shved, Vignesh Kothapalli, Yong Tang, marioecd

We are also grateful to all who filed issues or helped resolve them, asked and
answered questions, and were part of inspiring discussions.

Major Features
* Add `decode_avif` for AVIF image format support.
* Add `decode_jp3` for JPEG2000 image format support.
* Add `JPEG compression support for decoding TIFF images.
* Add audio spectrogram transform support.
* Add operations to trim/split/remix audio signals.
* Add Fade in/out audio augmentation support.
* Add frequency/time masking audio augmentation support.
* Add repeated field support in BigQuery API.

Thanks to our Contributors

This release contains contributions from many people:

Aleksey Vlasenko, Cheng Ren, Dio Gado, Ruhua Jiang, Yong Tang

We are also grateful to all who filed issues or helped resolve them, asked and
answered questions, and were part of inspiring discussions.

Major Features
* Add Python 3.8 support for Linux/macOS/Windows.
* Add Graph mode support for ArrowIOTensor.
* Add make_avro_dataset, parse_avro, and AvroRecordDataset
in tfio.experimental.columnar for Avro file format support.
* Add GeoTIFF support in decode_tiff as an extension.
* Add Video Capture support for macOS and Linux.
* Add decode_nv12 and decode_yuy2 in tfio.experimental.image
for colorspace conversion to RGB.
* Add decode_wav/encode_wav, decode_flac/encode_flac,
decode_vorbis/encode_vorbis(ogg) decode_mp3/encode_mp3,
decode_aac/encode_aac(mp4a) in tfio.audio for audio encoding and
decoding support.

Thanks to our Contributors

This release contains contributions from many people:

Aleksey Vlasenko, Ann Yan, Anthony Hsu, Bryan Cutler, Cheng Ren,
Florian Raudies, Keqiu Hu, Mark Daoust, Pei-Lun Liao, Pratik Dixit,
Yong Tang, Yuan Tang, Zou Xu, ashahab, marioecd

We are also grateful to all who filed issues or helped resolve them, asked and
answered questions, and were part of inspiring discussions.

Major Features
* Add Windows support for tensorflow-io.
* Add PostgreSQL server dataset support.
* Add MP3 format for audio dataset support.
* Add ArrowIOTensor with input from a pyarrow.Table.
* Add Numpy dataset support for numpy file and memory input.
* Add decode_avro/encode_avro for Avro serialization support.
* Prometheus dataset now outputs structured shapes.

Thanks to our Contributors

This release contains contributions from many people:

Bryan Cutler, DylanTallchiefGit, Kota Yamaguchi, Mark Daoust,
Suyash Kumar, Yong Tang, Yuan Tang

We are also grateful to all who filed issues or helped resolve them, asked and
answered questions, and were part of inspiring discussions.

Major Features
* Add Ogg audio format support.
* Add FLAC audio format support.
* Add HDR image format support.
* Extend HDF5 format support to all common types.

Thanks to our Contributors

This release contains contributions from many people:

Mark Daoust, Yong Tang, Yuan Tang, pshiko

We are also grateful to all who filed issues or helped resolve them, asked and
answered questions, and were part of inspiring discussions.