No abstract

In this paper we present TraceMove, a system to help novice animators create 2D, hand-drawn, character animation. The system and interface assists not only in sketching the character properly but also in animating it. A database of image frames, from recorded videos of humans performing various motions, is used to provide pose silhouette suggestions as a static pose hint to the users as they draw the character. The user can trace and draw over the generated suggestions to create the sketch of the pose. Then the sketch of the next frame of the animation being drawn is automatically generated by the system as a moving pose hint. In order to do this, the user marks the skeleton of the character in a single sketched pose, and a motion capture database is used to predict the skeleton for the subsequent frame. The sketched pose is then deformed according to the predicted skeleton pose. Furthermore, the sketch generated by the system for any frame can always be edited by the animator. This lets novice artists and animators generate hand-drawn 2D animated characters with minimal effort.

The prediction problem in any domain is very important to assess the prices and preferences among people. This issue varies for different kinds of data. Data may be nominal or ordinal, it may involve more categories or less. For any category to be considered by a machine learning algorithm, it needs to be encoded before any other operation can be further performed. There are various encoding schemes available like label encoding, count encoding and one hot encoding. This paper aims to understand the impact of various encoding schemes and the accuracy among the prediction problems of high cardinality categorical data. The paper also proposes an encoding scheme based on curated strings. The domain chosen for this purpose is predicting doctors’ fees in various cities having different profiles and qualification.

No abstract

No abstract