MAPL

[Journal/ OJCAS 2021]
Check out our work ANFIC: Image Compression Using Augmented Normalizing Flows.

1. ANFIC is an end-to-end learned image compression system based on Augmented Normalizing Flows (ANF). ANF is a new type of flow model, which stacks multiple variational autoencoders (VAE) for greater model expressiveness.

2. ANFIC advances further compression efficiency by stacking and extending hierarchically multiple VAE’s. The invertibility of ANF, together with our training strategies, enables ANFIC to support a wide range of quality levels without changing the encoding and decoding networks.

3. Extensive experimental results show that in terms of PSNR-RGB, ANFIC performs comparably to or better than the state-of-the-art learned image compression.

ANFIC is open source! The paper and the code are available at:
Paper:https://arxiv.org/pdf/2107.08470.pdf
Code: https://github.com/dororojames/ANFIC

https://www.linkedin.com/feed/update/urn:li:share:6963487445304635392?utm_source=linkedin_share&utm_medium=member_ios_link_share&utm_content=post

IEEE Circuits and Systems Society Rio Grande do Sul (CASS-RS) on LinkedIn: SBCCI TUTORIAL Wen Hsiao Peng (NCTU/Taiwan) Recent Advances in Learning-based SBCCI TUTORIAL Wen Hsiao Peng (NCTU/Taiwan) Recent Advances in Learning-based Image and Video Coding https://lnkd.in/dSmRp-kj...

[Conference/ ECCV 2022]
Check out our recent work CANF-VC: conditional augmented normalizing flows for video compression.

1. CANF-VC represents a new attempt that leverages the conditional ANF to learn a video generative model for conditional inter-frame coding.

2. CANF-VC extends the idea of conditional coding to motion coding, forming a purely conditional coding framework.

3. Extensive experimental results on commonly used datasets confirm the superiority of CANF-VC to the state-of-the-art methods.

CANF-VC is open source! The paper and the code are available at:

Paper: https://arxiv.org/abs/2207.05315
Code: https://github.com/NYCU-MAPL/CANF-VC

[Conference/CVPR2021]
Congratulations to Stanley (Yan-Cheng Huang), Julie (Yi-Hsin Chen), Cheng-You Lu, Hui-Po Wang on having one CVPR 2021 paper (Video Rescaling Networks with Joint Optimization Strategies for Downscaling and Upscaling) accepted for publication.

Project Page : http://mapl.nctu.edu.tw/MIMO_VRN/

[Conference/DCC2021]
Congratulations to Hector (Yung-Han Ho), Jerry (Guo-Lun Jin), Leo (Yun Liang) on having one DCC 2021 paper (A Dual-Critic Reinforcement Learning Framework for Frame-level Bit Allocation in HEVC/H.265) accepted for publication.

Project Page: http://mapl.nctu.edu.tw/RL_Rate_Control/

- CONFIRMED SPEAKER -
Prof. Wen-Hsiao Peng (NCTU, Taiwan)
Monday, October 19th
9:00 - 10:00 AM (GMT-3)

More information available at:
https://wp.ufpel.edu.br/dpvsa/wen-hsiao-peng/

Register to participate in the 4th IEEE DPVSA!
Registrations are open and are FREE OF CHARGE.
https://wp.ufpel.edu.br/dpvsa/registration/

VCIP 2020 - Paper Submission Deadline Approaching

JPEG - Call for Evidence on Learning-based Image Coding Technologies (JPEG AI) The JPEG Committee has launched the learning-based image coding activity, also referred to as JPEG AI. This activity aims to find evidence for image coding technologies that offer substantially better compression efficiency than available image codecs with models obtained from a large amount of visu...

JETCAS Special Issue on AICAS 2019 Contents The IEEE Journal on Emerging and Selected Topics in Circuits and Systems (JETCAS) proudly published its 2019 December issue to present recent technical advances in circuits and systems, ranging from cloud servers to edge devices and from system level to circuit level, to support the emerging artific...