In this paper, we present a novel example-based approach for cloning facial expressions of a source model to a target model while reflecting the characteristic features of the target model in the resulting animation. Our approach comprises three major parts: key-model construction, parameterization, and expression blending. We first present an effective scheme for constructing key-models. Given a set of source example key-models and their corresponding target key-models created by animators, we parameterize the target key-models using the source key-models and predefine the weight functions for the parameterized target key-models based on radial basis functions. In runtime, given an input model with some facial expression, we compute the parameter vector of the corresponding output model, to evaluate the weight values for the target key-models and obtain the output model by blending the target key-models with those weights. The resulting animation preserves the facial expressions of the input model as well as the characteristic features of the target model specified by animators. Our method is not only simple and accurate but also fast enough for various real-time applications such as video games or internet broadcasting.

No abstract

No abstract

We present language mechanisms for polymorphic, extensible records and their exact dual, polymorphic sums with extensible first-class cases. These features make it possible to easily extend existing code with new cases. In fact, such extensions do not require any changes to code that adheres to a particular programming style. Using that style, individual extensions can be written independently and later be composed to form larger components. These language mechanisms provide a solution to the expression problem.We study the proposed mechanisms in the context of an implicitly typed, purely functional language PolyR. We give a type system for the language and provide rules for a 2-phase transformationfirst into an explicitly typed λ-calculus with record polymorphism, and finally to efficient index-passing code. The transformations eliminate sums and cases by taking advantage of the duality with records.We implement a version of PolyR extended with imperative features and pattern matching-we call this language MLPolyR. Programs in MLPolyR require no type annotations-the implementation employs a reconstruction algorithm to infer all types. The compiler generates machine code (currently for PowerPC) and optimizes the representation of sums by eliminating closures generated by the dual construction.

We present language mechanisms for polymorphic, extensible records and their exact dual, polymorphic sums with extensible first-class cases. These features make it possible to easily extend existing code with new cases. In fact, such extensions do not require any changes to code that adheres to a particular programming style. Using that style, individual extensions can be written independently and later be composed to form larger components. These language mechanisms provide a solution to the expression problem. We study the proposed mechanisms in the context of an implicitly typed, purely functional language PolyR. We give a type system for the language and provide rules for a 2-phase transformationfirst into an explicitly typed λ-calculus with record polymorphism, and finally to efficient index-passing code. The transformations eliminate sums and cases by taking advantage of the duality with records. We implement a version of PolyR extended with imperative features and pattern matching-we call this language MLPolyR. Programs in MLPolyR require no type annotations-the implementation employs a reconstruction algorithm to infer all types. The compiler generates machine code (currently for PowerPC) and optimizes the representation of sums by eliminating closures generated by the dual construction. fun kv2kb kv = λv.'App(kv,[v]) fun kb2kv k b = withfresh(λxr.'Lam([xr],k b ('Var xr))) fun cvt app(e, e,kv) = let fun lc([],k b) = k b ([]) | lc(e:: e,k b) = pc(e, e,λ(v, v).k b (v:: v)) and pc(e, e,k b) = cvt(e,λv.lc(e,λ v.k b (v, v))) in pc(e, e,λ(v, v).'App(v,kv:: v)) end and cvt lam(x,e) = withfresh (λx k. 'Lam(x k :: x,cvt(e,kv2kb('Var x k)))) and cvt(e,k b) = match e with cases 'Con i ⇒ k b ('Con i) | 'Var x ⇒ k b ('Var x) | 'Lam(x,e)⇒ k b (cvt lam(x,e)) | 'App(e, e)⇒ cvt app(e, e,kb2kv(k b)) fun convert e = cvt lam([],e)