Kui Fan

Kui Fan

Auto-generated profile, based on publicly available information

Kui Fan

Shenzhen Graduate School, Peking University, Shenzhen, China

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
25 September 2019 in IEEE Transactions on Image Processing
Verified
Journal Article
Closed access
ZCZhuo Chen
Zhuo Chen
KFKui Fan
Kui Fan
SWShiqi Wang
Shiqi Wang
LDLingyu Duan
Lingyu Duan
WLWeisi Lin
Weisi Lin
AKAlex Chichung Kot
Alex Chichung Kot
The recent advances of hardware technology have made the intelligent analysis equipped at the front-end with deep learning more prevailing and practical. To better enable the intelligent sensing at the front-end, instead of compressing and transmitting visual signals or the ultimately utilized top-layer deep learning features, we propose to compactly represent and convey the intermediate-layer deep learning features with high generalization capability, to facilitate the collaborating approach between front and cloud ends. This strategy enables a good balance among the computational load, transmission load and the generalization ability for cloud servers when deploying the deep neural networks for large scale cloud based visual analysis. Moreover, the presented strategy also makes the standardization of deep feature coding more feasible and promising, as a series of tasks can simultaneously benefit from the transmitted intermediate layer features. We also present the results for evaluations of both lossless and lossy deep feature compression, which provide meaningful investigations and baselines for future research and standardization activities.

5 December 2018 in IEEE Transactions on Circuits and Systems for Video Technology
Verified
Journal Article
Closed access
KFKui Fan
Kui Fan
RWRonggang Wang
Ronggang Wang
GLGe Li
Ge Li
WGWen Gao
Wen Gao
The state-of-the-art high efficiency video coding (HEVC) standard provides a significant improvement relative to H.264/AVC, with almost 50% bitrate reduction. However, there are still requirements to further improve the coding performance of HEVC. In this paper, we reveal some problems in both intra and inter prediction by giving detailed observations, and solve them by further exploring spatial and temporal correlations. First, it is noteworthy that large boundary distortions in intra prediction blocks arise from the simplistic extrapolation operation with partial reference samples. To reduce these boundary distortions, we propose an enhanced prediction boundary filtering method based on generalized Gaussian distribution functions, and extend this method to merge prediction. Second, we combine intra and inter prediction to create a novel prediction method, in which spatial and temporal correlations are simultaneously considered. In addition, the existing merge mode can only provide accurate prediction results for simple movements, owing to the lack of motion estimation. To improve the flexibility, we optimize the existing merge mode by intensifying the temporal correlation with a simple motion search process. The experimental results show that the proposed methods achieve consistent bitrate savings, 1.3% for all intra, 1.5% for random access, 2.5% (up to 4.2%) for low delay P, and 1.2% for low delay B configuration. The encoding time only rises by 70% - 80%, and the decoding time increases negligibly.

8 March 2019 in IEEE Access
Verified
Journal Article
Open access
KFKui Fan
Kui Fan
RWRonggang Wang
Ronggang Wang
WLWeisi Lin
Weisi Lin
LDLing-Yu Duan
Ling-Yu Duan
WGWen Gao
Wen Gao

19 June 2020
Verified
Conference Paper
Open access
ZWZhenyu Wang
Zhenyu Wang
BHBingjie Han
Bingjie Han
RWRonggang Wang
Ronggang Wang
KFKui Fan
Kui Fan
WGWen Gao
Wen Gao
Compared with AVS2 and HEVC, the compression performance of the 3rd generation audio video coding standard (AVS3) has been greatly improved. For the preliminary application of this emerging coding standard, an efficient implementation of software-based decoder is necessary before the decoding chips are widely used. However, real-time decoding of UHD streams is still a big challenge for AVS3 standard. In this paper, we proposed a fast decoder design for AVS3, consisting of framework optimization, data structure and module optimizations, SIMD optimization, and parallel decoding. On the one hand, a non-redundant and cache-friendly decoding framework is proposed. In this framework, data structure and some key modules are optimized. On the other hand, SIMD instructions and frame-level parallelism are applied to fully accelerate decoding on a general-purpose processor. We also implemented an efficient AVS3 decoder (uAVS3d) in this paper. Experimental results show that, when decoding streams with bit depths of 10 and 8 on the i7--7800x processor, uAVS3d is 4.2 and 6.3 times faster than the AVS3 reference software with single-thread, respectively. uAVS3d has a speed-up of 6.2x when decoding with 17 threads comparing with single-thread on this processor, and can decode UHD streams in real-time. Additionally, uAVS3d has a decoding speed of 33.8 fps with single-thread for 1080p video on Huawei P10. Thus, uAVS3d can be used in most existing video applications on PC and mobile terminal device.