SIGGRAPH2011年论文 realtime performancebased facial animation

利用KINECT实现人脸表情动画实时演算和形状融合。利用kinect实时获取人脸的2D纹理和3D深度图，在电脑中利用表情的优先权重，来实现实时的动画
盘 Accumulated scans Accumulated Kinect Raw Depth Maps 3D Model Morphable model, manual Markup eneric Template Nonrigid ICP cumu Kinect Raw Images Userspecific Expressions Figure 3: Acquisition of user expressions for offine model build ing Aggregating multiple scans under slight head rotation reduces Examplebased noise and fills in missing data Facial Rigging )@自9 Generic Blendshapes we can use existing blendshape animations, that are ubiquitous in movie and game production, to define the dynamic expression pri 合@导 ors. The underlying hypothesis here is that the blendshape weights of a human facial animation sequence provide a sufficient level of Userspecific Blendshapes abstraction wo enable expression transfer between different charac ters. Finally, the output generated by our algorithm, a temporal se Figure 5: Offline preprocessing for building the userspecific ex quence of blendshape weights, can be directly imported into com pression model. Predefined example poses of the user with known mercial animation tools, thus facilitating integration into existing blendshape weights are scanned and registered to a template mesh production workflows to yield a set of userspecific expressions. An optimization solves for the userspecific blendshapes that maintain the semantics of a Acquisition Hardware. All input data is acquired using the generic blendshape model. The inset shows how manually selected Kinect system, i.e. no other hardware such as laser scanners is re feature correspondences guide the reconstruction of userspecific quired for userspecific model building. The Kinect supports si expressions multaneous capture of a 2D color image and a 3d depth map at 30 frames per second, based on invisible infrared projection (Figure 4 Essential benefits of this lowcost acquisition device include ease of essing step by adapting a generic blendshape model with a small deployment and sustained operability in a natural environment. The set of expressions performed by the user. These expressions are user is neither required to wear any physical markers or specialized captured with the Kinect prior to online tracking and reconstructed Makeup, nor is the per formance adversely affected by intrusive light using a morphable model combined with nonrigid alignment meth projections or clumsy hardware contraptions. However, these key ods. Figure 5 summarizes the different steps of our algorithm for dvantages come at the price of a substantial degradation in data building the facial expression model. We omit a detailed description quality compared to stateoftheart performance capture systems of previous methods that are integrated into our algorithm. Please based on markers and/or active lighting. Ensuring robust process refer to the cited papers for parameter settings and implementation ing given the low resolution and high noise levels of the input data details is the primary challenge that we address in this paper Data Capture. To customize the generic blendshape rig, we 2 Facial Expression Model record a predefined sequence of example expressions performe by the user. Since single depth maps acquired with the Kinect e central component of our tracking algorithm is a facial expres hibit high noise levels, we aggregate multiple scans over time using sion model that provides a lowdimensional representation of the the method described in [Weise et al. 2008](see Figure 3). The user user's expression space. We build this model in an offline prepro is asked to perform a slight head rotation while keeping the expres ion fixed (s ).B face to the uis rotational motion has the additional benefit of alleviating reconstruction bias introduced by the spatially fixed infrared dot pattern projected by the Kinect. We use the method of [viola and Jones 2001] to detect the face in the first frame of the acquisition and accumulate the acquired color images to obtain the skin texture using Poisson reconstruction [Perez et al. 2003 Figure 4: The Kinect simultaneously captures a 640 x 400 color Expression Reconstruction. We use the morphable model of Blanz and Vetter [1999] to represent the variations of different hu nage and corresponding depth map at 30 Hertz, computed via tri man faces in neutral expression. This linear PCa model is first reg angulation of an infrared projector and camera istered towards the recorded neutral pose to obtain a highquality normalized 0.1 max titi weight o frame rigid tracking mask Figure 6: The colored region on the left indicates the portion of the Figure 7: Robustly tracking the rigid motion of the face is crucial face used for rigid tracking. The graph on the rig ht illustrates how for expression reconstruction. Even with large occlusions and fast temporal filtering adapts to the speed of motion motion, we can reliably track the user's global pose the smoothed vector as weighted average in a window of size k as Template mesh that roughly matches the geometry of the users face We then warp this template to each of the recorded expressions us ing the nonrigid registration approach of [Li et al. 2009]. To im 飞、Y/t=3 prove registration accuracy, we incorporate additional texture con straints in the mouth and eye regions. For this purpose, we man ually mark features as illustrated in Figure 5. The integration of where ti denotes the vector at frame i3. The weights u'j are ned as constraints is straightforward and easily extends the frame of [Li et al. 2009] with positional constraints 们=e小Hm3x11,句lt4t=l∥ with a constant H that we empirically determine independently for rolation and translation based on the noise level of a salic pose. We Blendshape Reconstruction. We represent the dynamics of fa use a window size of k5 for all our experiments cial expressions using a generic blendshape rig based on Ekman's Facial Action Coding System(FACS)[1978]. To generate the full Scaling the lime scale with the Maximun variation in the temporal set of blendshapes of the user we employ examplebased facial rig window ensures that less averaging occurs for fast motion while ging as proposed by Li et al. [2010]. This method takes as input a highfrequency jitter is effectively removed from the estimated rigid generic blendshape model, the reconstructed example expressions pose(Figure 6, right). As shown in the video, this leads to a stable and approximate blendshape weights that specify the appropriate reconstruction when the user is perfectly still, while fast and jerky linear combinaLion of blendshapes for each expression. Since th motion can still be recovered accurately ser is asked to perform a fixed set of expressions, these weights are manually determined once and kept constant for all users. Given Nonrigld Tracking. Given the rigid pose, we now need to esti this data, examplebased facial rigging performs a gradientspace mate the blendshape weights that capture the dynamics of the facial optimization to reconstruct the set of userspecific blendshapes that expression of the recorded user. Our goal is Lo reproduce the users best reproduce the example expressions(Figure 5). We use th performance as closely as possible, while ensuring that the recon same generic blendshape model with m =39 blendshapes in all structed animation lies in the space of realistic human facial expres our examples sions. Since blendshape parameters are agnostic to realism and can easily produce nonsensical shapes, parameter fitting using geome 3 Realtime Tracking try and texture constraints alone will typically not produce satisfac tory results, in particular if the input data is corrupted by noise(see Figure &) Since human visual interpretation of facial imagery is The userspecific blendshape model defines a compact parameter highly sophisticated, even small tracking errors can quickly lead to space suitable for realtime tracking. We decouple the rigid from visually disturbing artifacts the nonrigid motion and directly estimate the rigid transform of the user's face before performing the optimization of blendshape 3.1 Statistical Model weights. We found that this decoupling not only simplifies the for mulation of the optimization, but also leads to improved robustness of the trackin We prevent unrealistic face poses by regularizing the blendshape weights with a dynamic expression prior computed from a set of existing blendshape animations A=(A1,., Au. Each anima Rigid Tracking. We align the reconstructed mesh of the previous tion A, is a sequence of blendshape weight vectors a; E R that frame with the acquired depth map of the current frame using ICP sample a continuous path in the mdimensional blendshape space with pointplane constraints. To stabilize the alignment we use We exploit temporal coherence of these paths by considering a win pI dow of n consecutive frames, yielding an effective prior for both the gion from the registration as this part of the face typically exhibits geometry and the motion of the tracked user the strongest deformations as illustrated in figure 7 this results in robust tracking even for large occlusions and extreme facial expres MAP Estimation. Let Di=Gi, li) be the input data at the sions. We also incorporate a temporal filter to account for the high current frame i consisting of a depth map Gi and a color im frequency flickering of the Kinect depth maps. The filter is based age I;. We want to infer from Di the most probable blendshape on a sliding window that dynamically adapts the smoothing coef weights xi E R for the current frame given the sequence X ficients in the spirit of the exponentially weighted moving average Xi1,.,Xin of the n previously reconstructed blendshape vec method [Roberts 1959] to reduce high frequency noise while avoid LOTS. Dropping the index i for notational brevity we formulate this ing disturbing temporal lags. We independently filter the translation inference problem as a maximum a posteriori(MAP)estimation vector and quaternion representation of the rotation. For a transla tion or quaternion vector ti at the current time frame i, we compute x= arg maxp(xD, Xn) where p(1 denotes the conditional probability. Using Bayes'rule we obtain x'= arg max p(Dx, Xn)p(x, Xn) 全會會 suning that d is conditionally independent of Xn given x, we can write x≈ arg max p(Dx)p(x,Xn likelihood Prior Distribution. To adequately capture the nonlinear structure of the dynamic expression space while still enabling realtime per formance, we represent the prior term p(x, Xn)as a Mixtures of Probabilistic Principal Component Analyzers(MPPCA)[Tipping and Bishop 1999b]. Probabilistic principal component analysis (PPCA)(see [Tipping and Bishop 1999a])defines the probabil ity density function of some observed data x E R by assuming that x is a linear function of a latent variable z e with s> t 詹圆會 1.e., x= Cz+H+e, 6) input data without prior with prior where z NN(O, I)is distributed according to a unit Gaussian, Figure 8: Without the animation prior tracking inaccuracies lead C E R is the matrix of principal components, u is the mean to visually disturbing selfintersections. Our solution significantly vector,and E w(o, a I) is a Gaussiandistributed noise variable reduces these artifacts. Even when tracking is not fully accurate as The probability density of x can then be written as in the bottom row, a plausible pose is reconstructed (x=N(xu, CC +oD USing this formulation, we define the prior in Equation 5 as a Likelihood Distribution. By assuming conditional indepen weighted combination of K gaussians dence, we can model the likelihood distribution in Equation 5 as the product P(Dx)=pG xp(Ix). The two factors capture the alignment of the blendshape Inodel with the acquired depth nap p(x,Xn)∑丌N(x,Xnk,CAC+n.(8) and texture image, respeclively. We represent the distribuLion of each likelihood term as a product of Gaussians, treating each vertex of the blendshape model independently with weights Tk. This representation can be interpreted as a reduceddimension Gaussian mixture model that attempts to model Let v be the number of vertices in the template mesh and B E the highdimensional animation data with locally linear manifolds y x m the blendshape matrix. Each column of b defines a blend modeled with ppca shape base mesh such that Bx generates the blendshape represen tation of the current pose. We denote with vi =(Bx)i the ith ver tex of the reconstructed mesh. The likelihood term p(Gx)models Learning the Prior. The unknown parameters in Equation 8 are a geometric registration in the spirit of nonrigid ICP by assuming the means uk, the covariance matrixes CK Ck, the noise parame a gaussian distribution of the pervertex pointplane distances ters k, and the relative weights Tk of each PPCa in the mixture Inode. We learn Chese paraineters using the expectation maxiiniza tion(eM)algorithm based on the given blendshape animation se A. To increase the robustness of these computations, we P(Gx)=II n(viv) exp(一 ), 21(2丌geo )号 estimate the MPPCa in a latent space of the animation sequences L using principal component analysis. By keeping 99% of the to tal variance we can reduce the dimensionality of the training data where ni is the surface normal at Vi, and vi is the corresponding by twothirds allowing a more stable learning phase with the em closest point in the depth map g algorithm. Equation can thus be rewritten as The likelihood term p(Ix)models texture registration. Since we acquire the users face texture when building the facial expression p(x,Xn)=∑丌N(x,X1Pk+,PMP model(Figure 3), we can integrate modelbased optical How con straints [Decarlo and Metaxas 2000], hy formulating the likelihood function using pervertex Gaussian distributions as where M =(Ck Ck +okn is the covariance matrix in the latent space, Pis the principal component matrix, and u the mean vector Since the em algorithm converges to local minima. we run the al p(x)=l2ra2 exp( 2a2 2一),(11) gorithm 50 times with random initialization to improve the learnin accuracy. We use 20 Gaussians to model the prior distribution and we use onethird of the latent space dimension for the ppca d where pi is the projection of vi into the image 1,vli is the gradient mension. More details on the implementation of the em algorithm of I at pi, and pi is the corresponding point in the rendered texture can be found in [Mclachlan and Krishnan 1996 image 3.2 Optimization In order to solve the map problem as defined by equation 5 we minimize the negative logarithm,1.e 堡窗回 x'=arg minInP(Gx)Inp(Ix)Inp(x, Xn).(12) Discarding constants, we write 會@最 x= arg min Egeo+ Eim prior (13) mplx, xn) ∑ln3(v,v)‖,and (15) 量量 ∑l The parameters o geo and im model the noise level of the data that conTrols the emphasis of the geonetry and inage likelihood terMs relative to the prior term. Since our system provides realtime feed back, we can experimentally determine suitable values that achieve stable tracking performance. For all our results we use the same settings geo= 1 and oim =0.45 The optimization of equation 13 can be performed efficiently using an iterative gradient solver, since the gradients can be computed tracked analytically(see the derivations in the Appendix). In addition, we input data virtual avatars expression model precompute the inverse covariance matrices and the determinants of the MPPCa during the offine learning phase. We use a gradi ent projection algorithm based on the limited memory BFGS solver [Lu et al. 1994] in order to enforce that the blendshape weights are between 0 and 1. The algorithm converges in less that 6 iterations as we can use an efficient warm starting with the previous solution We then update the closest point correspondences in Figec and Fim 每最 and recompute the MAP estimation. We found that 3 iterations of this outer loop are sufficient for convergence blendshape base meshes 4 Results Figure 9: The user's facial expressions are reconstructed and mapped to different target characters in realtime, enabling inter We present results of our realtime performance capture and anima active animations and virtual conversations controlled by the per tion system and illustrate potential applications. The output of the formance of the tracked user. The smile on the green character's tracking optimization is a continuous stream of blendshape weight base mesh gives it a happy countenance for the entire animation vectors x; that drive the digital character. Please refer to the ac companying video to beller appreciate the lacial dynamics of the animated characters and the robustness of the tracking, figures 1 Statistics. We use 15 userspecific expressions to reconstruct 39 and 9 illustrates how our system can be applied in interactive appli blendshapes for the facial expression model. Manual markup of cations where the user controls a digital avatar in realtime. blend texture constraints for the initial offline model building requires ap chape weights can be transmitted in realtime to enable virtual en counters in cyberspace. Since the blendshape representation facili proximately 2 minutes per expression. Computing the expression model given the user input takes less than 10 minutes. We pre tates animation transfer, the avatar can either be a digital representa compute the gaussian mixture model that defines the dynamic ex tion of the user himself or a different humanoid character, assuming pression prior from a total of 9, 500 animation frames generated on compatible expression space the generic template model by an animation artist. Depending on the size of the temporal window, these computations take between While we build the userspecific blendshape model primarily for 10 and 20 minutes realtime tracking, our technique offers a simple way to create per sonalized blendshape rigs that can be used in traditional animation Our online system achieves sustained framerates of 20 Hertz with tools. Since the Kinect is the only acquisition device required, gen a latency below 150 ms. Data acquisition, preprocessing, rigid reg erating facial rigs becomes accessible for nonprofessional users istration, and display take less than 5 ms. Nonrigid registration 匾应 input dala expression model Figure ll: Difficult tracking configurations. Right: despite the occlusions by the hands, our algorithm successfully tracks the rigid input data geometry only texture only geometry+texture motion and the expression of the user: Left: with more occlusion or very fast motion, tracking can fail Figure 10: The combination of geometric and texturebased tration is essential for realtime tracking To isolate the effects individual components, no animation prior is used in this example tracked model more closely matches the performing user. What the prior achieves in any case is that the reconstructed pose is plau sible, even if not necessarily close to the input geometrically(see including constraint setup and gradient uptimization require 45 ms also Figure 8). We argue that this is typically much more tolerable per frame. All timing measurements have been done on a Intel I7 than generating unnatural or even physically impossible poses that 2. 8 Ghz with 8 GBytes of main memory and a ATI Radeon HD 4850 could severely degrade the visual perception of the avatar. In addi graphics card tion,our approach is scalable in the sense that if the reconstructed animation does not well represent certain expressions of the user, 5 Evaluation we can manually correct the sequence using standard blendshape animation tools and add the corrected sequence to the training data We focus our evaluation on the integration of 2D and 3D input data set. This allows to successively improve the animation prior in a and the effect of animation training data. we also comment on bootstrapping manner. For the temporal window Xn used in the limitations and draw backs of our approach animation prior, we found a window size of3< n 5 to yield good results in general. Longer temporal spans raise the dimen signality and lead to increased temporal smoothing. If the window Geometry and texture figure 10 evaluates the interplay be is too small, temporal coherence is reduced and discontinuities in tween the geometry and texture information acquired with the the tracking data can lead to artifacts Kinect Tracking purely based on ge Limitations. The resolution of the acquisition system lir ometry as proposed in [Weise amount of geometric and motion detail that can be tracked for each et al. 2009] is not successful due user, hence slight differences in expressions will not be captured to the high noise level of the adequately. This limitation is aggravated by the wideangle lens Kinect data. Integrat del of the Kinect installed to enable fullbody capture which confines based optical flow constraints the face region to about 160x 160 pixels or less than 10%o of the reduces temporal jitter and sta Lotal image area. As a resull, our systen cannot recover sInallscal WEise et al. 200 bilizes the reconstruction, In our wrinkles or very subtle movements. We also currently do not model riments, only the co cyes, teeth, tongue, or hair tion of both modalities yielded satisfactory results. Compared to purely imagebased tracking as e. g in [Chai et al. 2003, direct ac In our current implementation, we require user support during pre cess to 3D geometry offers two main benefits: We can significant processing in the form of manual markup of lip and eye features improve the robustness of the rigid pose estimation in particular for lo register the generic TeMplate with the recorded training poses nonfrontal views(see also Figure 7). In addition, the expression (see Figure 5). In future work, we want to explore the potential of template mesh generated during preprocessing much more closely generic active appearance models similar to [Cootes et al. 2001]to matches the geometry of the user, which further improves track automate this step of the offline processing pipeline as well ing accuracy. Figure II shows difficult tracking configurations and provides an indication of the limits of our algorithm While offering many advantages as discussed in Section 1. 2, the blendshape representation also has an inherent limitation: The number of blendshapes is a tradeoff between ex pressiveness of the Animation Prior Figure 12 studies the effectiveness of our prob model and suitability for tracking. Too few blendshapes may re abilistic tracking algorithm when varying the amount of training sult in user expressions that cannot be represented adequately by data used for the reconstruction. The figure illustrates that if the the pose space of the model. Introducing additional blendshapes training data does not contain any sequences that are sufficiently to the rig can circumvent this problem, but too many blendshapes close to the captured performance, the reconstruction can differ may result in a different issue: Since blendshapes may become ap substantially from the acquired data. with more training data, the proximately linearly dependent, there might not be a unique set of 堡回园园回 Acknowledgements. We are grateful to Lee PerrySmith for ding the fa template, Dan burke for sculpting the CG characters, and Cesar Bravo, Steven Mclellan, David rodrigues, and volker Helzle for the animations. We thank Gabriele Fanelli for our valuable discussions, duygu Ceylan and Mario Deuss for being actors, and Yuliy schwarzburg for proof reading the paper. This research is supporled by Swiss National Science Foundation grant 20PA2IL129607 Appendi We derive the gradients for the optimization of e 13.Th 鱼霞會會會會 energy terms for geometry registration geo and optical fow Lim can both be written in the form input data 100% ∫(ⅹ I Axb Figure 12: Effect of different amounts of training data on the per hence the gradients can easily be computed analytically as formance of the tracking algorithm. We successively delete blend shapes from the input animation sequences, which removes entire af (x) A(Axb) (18) portions of the expression space. With only 25%of the blendshapes dx in the training data the expressions are not reconstructed correctly The prior term is of the form blendshape weights for a given expression. This can potentially re ∑丌N(x,xnk,∑k) sult in unstable tracking due to overfitting of the noisy data. While the prior prevents this instability, a larger number of blendshapes re where 2k is the covariance matri> The Gaussians quires a larger training database and negatively affects performance W(x, Xnu:, 2k)model the combined distribution of the current blendshape vector x E R and the n previous vectors Xn 6 Conclusion hence the k are matrices of dimension(n +i)m m+1)m Since we are only interested in the gradient with respect to x, we We have demonstrated that highquality performancedriven fa can discard all components that do not depend on this variable. We cial animation in realtime is possible even with a lowcoSt, non split the mean vectors as uk =(uk, uk), corresponding to x and ntrusive, markerless acquisition system. We show the potential of Xn, respeclively. We can write the inverse of 2kas our system for applications in human interaction, live virtual TV shows, and computer gaming Ak B (m×m)(m×n:m) (20) Robust realtime tracking is achieved by building suitable user C (m×m)(mm×nm) specific blendshape models and exploiting the different character islics of the acquired 2D image and 3D deplh Inap data for regis tration. We found that learning the dynamic expression space from with Bk=Ck. We then obtain for the gradient of the prior energy existing animations is essential. Combining these animation priors ith effective geometry and texture registration in a single MAP es aPrior timation is our key contribution to achieve robust tracking even for highly noisy input data. While foreseeable technical advances in ∑k=1xkN(x,Xn4k,∑k)(x)rAk+(Xn12)Ck acquisition hardware will certainly improve data quality in coming ars, numerous future applications, e.g. in multipeople tracking, 乙k=17kN(x,Xnux,∑k acquisition with mobile devices, or performance capture in diffi The complete gradient is the sum of the three energy gradients de cult lighting conditions, will produce even worse data and will thus rived above systematic framework for addressing these challenging prublellf 9 put even higher demands on robustness. Our algorithm provides 9(x)=dEgeo a+ ax ae prior ax ax (22) We believe that our system enables a variety of new applications and can be the basis for substantial followup research. We cur rently focus on facial acquisition and ignore other important as pects of human communication, such as hand gestures, which pose interesting technical challenges due to complex occlusion patterns Enhancing the tracking performance using realtime speech analy sis, or integrating secondary effects such as simulation of hair are further areas of future research that could help increase the realism of the generated virtual performances. More fundamentally, being able to deploy our system at a massive scale can enable interesting new research in human communication and paves the way for new interaction metaphors in performancebased game play References LIH, ROIVAINEN P. and FORCHEiMERR. 1993. 3d motion estimation in modelbased facial image coding. PAM15,545 aleXaNDER,O. ROGERS. M. LAMBETH W. CHANG. M AND DEBEVEC, P. 2009. The digital emily project: photoreal facial modeling and animation. ACM/GGRAPH 2009 Courses LI.H.. ADAMSB. GUIBAS. L.J. AND PAULY.M. 2009. Robust singleview geometry and motion reconstruction. ACM Trans BEELER, T, BICKEL, B, BEARDSLEY, P, SUMNER, B, AND Graph.28,175:1175:10. GROSS,M. 2010. Highquality singleshot capture of facial geometry. ACM Trans. Graph. 29, 40: 140: 9 LI, H, WEISE, T, AND PAULY, M. 2010. Examplebased facial rigging. ACM Trans. Graph. 29, 32: 132: 6 BLACK, M.J., AND YACOOB, Y. 1995. Tracking and recognizing rigid and nonrigid facial motions using local parametric models LiN, I.C, AND OUHYOUNG, M. 2005. Mirror mocap: Automatic of image motion In Iccv.37438 and efficient capture of dense 3d facial motion parameters from video. The visual Computer 21, 6, 355372 BLANZ. V, AND VETTER, T.1999. A morphable model for the LOU, H, AND CHAL, J. 2010. Examplebased human motion synthesis of 3d faces. In Proc. SIGGRAPH 99 denoising IEEE Trans. on visualization and Computer graphics BORShUKOV.G. PipoNI D. larseN. O. LEWIS.J.P. AND 16,870879 TEMPELAARLIETZ, C. 2005. Universal capture  imagebased facial animation for the matrix reloaded. In siggraph 2005 LU, P, NOCEDAL, J, ZHU, C, BYRD, R. H, AND BYRD,R. H 1994. A limitedmemory algorithm for bound constrained opti Courses mization. SIAM Journal on Scientific Computing BRADLEY. D. HEIDRICH. W. POPA. T. AND SHEFFER. A MA. W.C. HaWkins. T. Peers. P. CHABERT. C.F.. WEISS 2010. High resolution passive facial performance capture. ACM M., AND DEBEVEC, P. 2007. Rapid acquisition of specular and Tran. Graph.29,41:141:10 diffuse normal maps from polarized spherical gradient illumina CHAI,J.ⅹ.,ⅹIAO,J., AND HODGINS,J.2003.Ⅴ ISion based tion In EUROGRAPHICS Symposium on Rendering control of 3d facial animation In sca MCLACIILAN, G. J AND KRISIINAN, T. 1996. The EM Algo CHUANG, E, AND BREGLER, C. 2002. Performance driven facial rithm and Extensions. WileyInterscience animation using blendshape interpolation. Tech. rep, Stanford PEREZ, P, GANGNET, M, AND BLAKL, A. 2003. Poisson image University editing. ACM Trans. Graph. 22, 313318 COOTES. T. EDWARDS. G. AND TAYLOR. C. 2001. Active PIGHIN. F. AND LEWIS. J. P. 2006. Performancedriven facial appearance models. PAM/ 23, 681685 imation In ACm siggraPh 2006 Courses COVELL, M. 1996. Eigenpoints: Controlpoint location using PIGHIN, F, SZELISKI,R, AND SALESIN, D. 1999. Resynthesiz principle component analyses. In FG 96 ing facial animation through 3d modelbased tracking. ICCV 1, 143150. DECARLO, D. AND METAXAS, D. 1996. The integration of optical flow and deformable models with applications to human ROBERTS, S. 1959. Control chart tests based on geometric moving face shape and motion estimation. In CVPR averages. In Tech nometrics. 239250 DECARLO, D, AND METAXAS, D. 2000. Optical flow constraints TIPPING, M.E., AND BISHOP, C.M. 1999. Probabilistic principal on deformable models with applications to face tracking. IJCV component analysis. Journal of the royal statistical society 38.99127 Series b EKMAN, P, AND FRIESEN, W. 1978. Facial Action Coding Sys TIPPING, M.E., AND BISHOP, C. M. 19g9. Mixtures of proba tem: A Technique for the Measurement of Facial Movement. bilistic principal component analyzers. Neural Computation 11 Consulting Psychologists Press VIOLA, P, AND JONES, M. 2001. Rapid object detection using a ESSA, I. BASU, S, DARRELL, T, AND PENTLAND, A. 1996 boosted cascade of simple features. In CVPR Modeling, tracking and interactive animation of faces and heads using input from video. In Proc. Computer Animation WEISE T.FIBE. B. AND GOOL I.v. 2008. Accurate and robust registration for inhand modeling In CVPR FURUKAWA, Y, AND PONCE, J. 2009. Dense 3d motion capture WEISE. T. GOOL.L.V. AND PAULY. M. 2009. Face/off for human faces. In CVPR Live facial puppetry. In SCa GROCHOW, K. MARTIN. S. L. HERTZMANNA. AND WILLIAMS. L. 1990. Performancedriven facial animation In POPOVIC, Z. 2004. Sty lebased inverse kinematics. ACM Trans Comp Graph( Proc. SIGGRAPH 90) Graph.23,522531 WilSON. C.A. GHOSH.A. Peers P. CHiANG.Y. BUSCH GuenteR. B. GRIMM. C. WOOD. D. MALVAR H. AND J, AND DEBEVEC, P. 2010. Temporal upsampling of per PIGHIN, F. 1993. Making faces. IEEE Computer Graphics formance geometry using photometric alignment ACM Trans and Applications 13, 68 Graph.29,17:117:11 IKEMOTO. L. ARIKANO. AND FORSYTH. D. 2009. General ZHANG, S, AND HUANG, P. 2004. Highresolution, realLine 3d izing motion edits with gaussian processes. ACM Trans. graph shape acquisition. In CvPR Workshop 28.1:11:12 ZHANG, L, SNAVELY, N, CURLESs, B, AND SEITZ, S. M LAU, M. CHAL, J, XU, Y.Q., AND SHUM,H Y. 2007. Face 2004. Spacetime faces: high resolution capture for modelin poser: interactive modeling of 3d facial expressions using model and animation. ACM Trans. Graph. 23, 548558 riors. In sCa

20131231

20130419

20121201
 3.18MB
Realtime PerformanceBased Facial Animation.pdf
20140518基于kinect的人脸重建，取出纹理和噪点，将建模用于后续点对点的时事交流
 539KB
3D Shape Regression for Realtime Facial Animation
201408213D Shape Regression for Realtime Facial Animation
 6.43MB
SIGGRAPH会议论文精选
20100304SIGGRAPH2002年会议论文精选，SIGGRAPH2002年会议论文精选
 1.16MB
论文报告：Realtime Skeletal Skinning with Optimized Centers of Rotation
20170407对于论文Realtime Skeletal Skinning with Optimized Centers of Rotation和Efficient Dynamic Skinning with L
 3.2MB
SIGGRAPH 99 Course OpenGL 课程讲义
20111116SIGGRAPH 99 Course Notes OpenGL SIGGRAPH 99 Course OpenGL 讲义 打包，源地址： ftp://ftp.sgi.com/opengl/contri
 33KB
Exp_AnchoredMomentumCandle  MetaTrader 5EA.zip
20190913Exp_AnchoredMomentumCandle EA交易是基于AnchoredMomentumCandle 指标信号的.
 3.71MB
孢子动画实现技术分析 Sporeanimsiggraph08.pdf
20170331孢子，Spore, Animation 动画
 28.79MB
Subdivision for Modeling and Animation
20110926a course on surface subdivision algorithms presented by experts in the SIGGRAPH course session.
 1.88MB
3DDisplay_USCICT_SIGGRAPH2007[1]
201308023DDisplay技术，通过对实体的三维数据竞相整合，以动态的方式扫描投影，显示三维效果。
 21.57MB
siggraph2019 2 Navigating Intrinsic Triangulations.pdf
20191016siggraph2019 2 Navigating Intrinsic Triangulations.pdf
 54.43MB
Stochastic ScreenSpace Reflections
20170816Talk by Tomasz Stachowiak at SIGGRAPH 2015 in the Advances in Realtime Rendering course
 2.25MB
Geometric Data Structures for Computer GraphicsSiggraph 2003 Tutorial 16.pdf
20200322In recent years, methods from computational geometry have been widely adopted by the computer graphi
 5KB
计算机视觉 第一次试验
20190104计算机视觉第一次大作业主要代码，实验一 :混合图像实验目标是编写一个图像滤波函数，并用它基于Oliva、Torralba和Schyns在SIGGRAPH 2006发表的题为“Hybrid images
 665KB
2D流体算法（C++实现）
20190102SIGGRAPH论文 Stam, J.: Stable fluids. In Proceedings of SIGGRAPH. (1999), pp. 121128. 文中的算法。
 354KB
Stable Fluids
20180317关于流体模拟的一片经典论文。 animation of fluids, NavierStokes, stable solvers, implicit elliptic PDE solvers, in
 1.43MB
Visual Simulation of Smoke
20180105siggraph经典论文，烟雾模拟，可以扩展到火焰的模拟。代码量不大，易于模拟。
 29.59MB
给黑白图像自动上色的win程序（2）
20190518可以给黑白图像上彩色的程序（win下），由《siggraph2016_colorizationmaster》中的 colornet.t7 模型改编而来（修正了前面的一个错误，已经达到和《siggra
 1.52MB
deepsketch2face, 在 SIGGRAPH 2017中，演示"DeepSketch2Face"的演示和数据库.zip
20190917deepsketch2face, 在 SIGGRAPH 2017中，演示"DeepSketch2Face"的演示和数据库 DeepSketch2Face注意：数据库和带有双线性编码的演示将在不久的将来
 2.6MB
混合图像Exp1_Hybrid_Images
20190912实验目标是编写一个图像滤波函数，并用它基于Oliva、Torralba和Schyns在SIGGRAPH 2006发表的题为“Hybrid images”的论文的简化版本创建混合图像。混合图像是静态图像
 25.97MB
浙江大学研究生课程计算机视觉课件
20090529浙江大学计算机学院，世界图形学大会siggraph论文第一作者王锐博士的授课课件。
 7.1MB
图像平滑L0
20150122SIGGRAPH2012中的论文，用于图像平滑，平滑同时能够保持很好的目标边界
 11.61MB
Grabcut算法源码
20100804SIggraph 2004 论文“GrabCut” — Interactive Foreground Extraction using Iterated Graph Cuts 的源代码。
MySQL基础入门视频课程
20181127本课程从零开始，以通俗易懂的方式讲解MySQL技术，手把手教你掌握每一个知识点。课程中使用的所有英文单词都会逐一查询并记录，真正做到零基础入门学习，适合初学者的教程！ 课程内容包括： 1.MySQL简介、安装MySQL 2.查询操作 3.聚合函数和分组统计 4.更新操作 5.表和库的管理 6.约束 7.用户和权限管理 8.事务处理 教学全程采用笔记+代码案例的形式讲解，通俗易懂！！！

学院
python基础快速处理Excel
python基础快速处理Excel

博客
代码执行
代码执行

下载
嵌入式系统/ARM技术中的基于FPGA的SOPC系统DAB发射端硬件实现
嵌入式系统/ARM技术中的基于FPGA的SOPC系统DAB发射端硬件实现

学院
3天计算机视觉精进训练营
3天计算机视觉精进训练营

博客
分布式理论：一致性协议 3PC
分布式理论：一致性协议 3PC

学院
Java微信小程序校园图书商城 大学生毕业设计教学视频
Java微信小程序校园图书商城 大学生毕业设计教学视频

博客
今天的c复习截图
今天的c复习截图

博客
20201020
20201020

博客
【ALGO】Leetcode 91.解码方法
【ALGO】Leetcode 91.解码方法

下载
php cookie用户登录的详解及实例代码
php cookie用户登录的详解及实例代码

博客
20201020
20201020

学院
系统学习Linux命令
系统学习Linux命令

博客
waterdrop导出hdfs数据到clickhouse(text,csv,json)
waterdrop导出hdfs数据到clickhouse(text,csv,json)

博客
实例QT程序 —— Qt单例不规则界面程序
实例QT程序 —— Qt单例不规则界面程序

下载
详解Yii2 定制表单输入字段的标签和样式
详解Yii2 定制表单输入字段的标签和样式

学院
Java小白学习方法指南！避开思维惯性！让学习事半功倍
Java小白学习方法指南！避开思维惯性！让学习事半功倍

下载
地铁车站弱电系统集中UPS供电分析
地铁车站弱电系统集中UPS供电分析

下载
注塑车间自动信息化管理系统
注塑车间自动信息化管理系统

学院
PHP微信小程序服装购物商城大学生毕业设计教学视频
PHP微信小程序服装购物商城大学生毕业设计教学视频

下载
数据转换/信号处理中的高精度数模转换器的温漂研究
数据转换/信号处理中的高精度数模转换器的温漂研究

博客
命令执行漏洞
命令执行漏洞

学院
Spring AOP从认识到源码
Spring AOP从认识到源码

下载
台达自动化产品在线路板封胶机中的应用
台达自动化产品在线路板封胶机中的应用

学院
Web全栈之成为编程小牛的十八般武艺秘籍
Web全栈之成为编程小牛的十八般武艺秘籍

下载
嵌入式系统/ARM技术中的基于国产 CPU 的嵌入式医疗电子无线网络设计
嵌入式系统/ARM技术中的基于国产 CPU 的嵌入式医疗电子无线网络设计

下载
嵌入式系统/ARM技术中的嵌入式现场总线通信网关的设计与实现（一）
嵌入式系统/ARM技术中的嵌入式现场总线通信网关的设计与实现（一）

博客
Pycharm中Flask的调试模式
Pycharm中Flask的调试模式

学院
JavaWeb服装购物商城毕业设计 大学生毕业设计教学视频
JavaWeb服装购物商城毕业设计 大学生毕业设计教学视频

学院
【数据分析】量化交易
【数据分析】量化交易

学院
2020年_Python_Django基础+实战全套视频【千锋】
2020年_Python_Django基础+实战全套视频【千锋】