Automatic motion-guided video stylization and personalization

Cao, Chen; Chen, Shifeng; Wei, Zhang; Tang, Xiaoou

doi:10.1145/2072298.2071933

Cited by 9 publications

(5 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Cao et al [4] adapted the idea of image analogies [8] to achieve the multi-style video stylization system. This method divides the first stylized video frame into many blocks and then propagates these patches according to the optical flow field, which is optimized by occlusion detection and bilateral diffusion.…”

Section: Related Workmentioning

confidence: 99%

“…Furthermore, this method needs to calculate the distance between the overlap regions of the adjacent blocks to determine whether re-rendering these regions is necessary. Unlike the algorithm in [4], we neither need to optimize the optical flow field nor calculate the correlation between blocks to ensure temporal coherence. We only need to detect the distorted texture layer area during texture advection and perform limited texture in-painting.…”

Section: Related Workmentioning

confidence: 99%

“…Constrained by the optical flow field, the texture layer is advected and combined with the video frames to achieve the final results. The goal of our method is similar to [4] since we both aim to achieve video personalization with chosen styles. To achieve better effects of customized styles, we consider the texture orientation during synthesis.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Multi-style video stylization based on texture advection

Tang

Zhang

Shi

et al. 2015

Sci. China Inf. Sci.

View full text Add to dashboard Cite

Artistic video stylization, which is widely used in multimedia entertainment, transforms a given video into different artistic styles. Most of the existing video stylization algorithms can simulate single or limited video artistic styles. Although some algorithms can achieve multi-style video processing, these algorithms are complex and difficult to implement. To solve this problem, we propose a multi-styled video stylization algorithm based on texture advection, where different artistic styles are synthesized and transferred from user-specified texture samples of desired styles. We use the direction field-guided texture synthesis to compute the texture layer that represents the artistic style. Painterly directional video styles are simulated competently by the orientation changes in the synthesized anisotropic textures. There appeared local distorted region of the texture layer during texture advection under the optical flow field. To address this issue, we propose the texture inpaint to synthesize the limited distorted region and make the stylized video temporally coherent. We also accelerate the video stylization by using the CUDA parallel computing framework that parallelly computes the morphological operations used for video abstraction. Finally, we produce stylized videos of multiple artistic styles with satisfactory experimental results, including the styles of oil painting, watercolor painting and stylized lines drawing.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Multi-style video stylization based on texture advection

Tang

Zhang

Shi

et al. 2015

Sci. China Inf. Sci.

View full text Add to dashboard Cite

show abstract

“…However, none of these is fully satisfactory, causing either blending, sliding or 3D distortion artifacts. Cao et al [2011] use a comparable approach to stylize videos by relying on optical flow. Hashimoto et al [2003] and Haro [2003] applied the Image Analogies method to low-resolution video sequences, using motion estimation to increase temporal coherence.…”

Section: Synthesis For Animationmentioning

confidence: 99%

Stylizing animation by example

Bénard

Cole²,

Kass³

et al. 2013

ACM Trans. Graph.

View full text Add to dashboard Cite

Figure 1: Frame 20 (right) to 34 (left) -top row: stylized animation; bottom row: input shaded images. The two extreme frames are keyframes painted by an artist. Our algorithm synthesized the in-between frames. Note the important variations in terms of shape and appearance (texture and color) during this sequence. The accompanying video illustrates the temporal behavior. AbstractSkilled artists, using traditional media or modern computer painting tools, can create a variety of expressive styles that are very appealing in still images, but have been unsuitable for animation. The key difficulty is that existing techniques lack adequate temporal coherence to animate these styles effectively. Here we augment the range of practical animation styles by extending the guided texture synthesis method of Image Analogies [Hertzmann et al. 2001] to create temporally coherent animation sequences. To make the method art directable, we allow artists to paint portions of keyframes that are used as constraints. The in-betweens calculated by our method maintain stylistic continuity and yet change no more than necessary over time.

show abstract

“…Traditional researches in the NPR area are usually validated by subjective/visual experiments, as in the research of Kang et al (2015), Wenhua et al (2015), Zhang et al (2013), Baugh and Kokaram (2010), Benard et al (2012), Borawski (2014), Cao et al (2011), Chen et al (2012), Gangopadhyay et al (2016) and Wang et al (2017). However, purely subjective/visual evaluation, without the use of external objective criteria and the lack of a representative sample of participants, weakens the statistical validity of the analyses performed.…”

Section: Introductionmentioning

confidence: 99%

Systematic Evaluation of Image-Based NPR Applied to Video

Barros¹,

Queiroz²,

Gomes³

2020

Journal of Computer Science

View full text Add to dashboard Cite

In this article, it was studied the effect of state-of-the-art approaches for Non-Photorealistic Rendering (NPR), originally developed for static images, when applied to videos. Six criteria were objectively evaluated (using a 5-point Likert scales): Simplicity, content preservation, resemblance to the original content, sharpness, temporal coherence and subjective satisfaction. Sixty participants and a dataset of 30 videos of different genres were used in the experiments. Each participant voted 15 pairwise video comparisons. Three approaches were considered: Anisotropic Kuwahara, Coherent Enhancement and Extended Difference of Gaussians (XDoG). The results of the study, under 99% statistical confidence, indicated that the Anisotropic Kuwahara approach achieved the best results for most of the criteria. Coherent Enhancement and XDoG approaches came next. These results indicate that the selected approaches were considered satisfactory, when applied to video, for most considered criteria.

show abstract

Automatic motion-guided video stylization and personalization

Cited by 9 publications

References 12 publications

Multi-style video stylization based on texture advection

Multi-style video stylization based on texture advection

Stylizing animation by example

Systematic Evaluation of Image-Based NPR Applied to Video

Contact Info

Product

Resources

About