Search

Advancements on text-to-image synthesis generate remarkable images from textual descriptions. However, these methods are designed to generate only one object with varying attributes. They face difficulties with complex descriptions having multiple arbitrary objects since it would require information on the placement and sizes of each object in the image. Recently, a method that infers object layouts from scene graphs has been proposed as a solution to this problem. However, their method uses only object labels in describing the layout, which fail to capture the appearance of some objects. Moreover, their model is biased towards generating rectangular shaped objects in the absence of ground-truth masks. In this paper, we propose an object encoding module to capture object features and use it as additional information to the image generation network. We also introduce a graph-cuts based segmentation method that can infer the masks of objects from bounding boxes to better model object shapes. Our method produces more discernible images with more realistic shapes as compared to the images generated by the current state-of-the-art method.

Researches dealing with the task of style transfer have focused and produced results that only transfers the style of a reference image to the entirety of another image. As this domain would be beneficial in digital art, it would be preferable for such algorithms to support a more flexible style transfer, such that the style would be applied only on a specific portion of an image. We propose a framework that can selectively apply a given style onto an object using only 4 user-defined points. Our approach combines a style transfer module and an object segmentation module to synthesize the stylized image. As the ultimate goal is to develop an artistic application tool, we also introduce a method that makes use of specific filters to preserve certain characteristics of an image, such as its high frequency components. Experiments show that our proposed method is able to produce pleasing results with minimal effort while also being able to handle more complicated tasks such as the application of multiple reference style images onto different objects in an image.

Gender-specific classifiers are shown to outperform general classifiers. In calibrated experiments designed to demonstrate this, two sets of data were used to build male-specific and female-specific classifiers. The first dataset is used to predict vowel phonemes based on speech signals, and the second dataset is used to predict negative emotions based on brainwave (EEG) signals. A Multi-Layered-Perceptron (MLP) is first trained as a general classifier, where all data from both male and female users are combined. This general classifier recognizes vowel phonemes with a baseline accuracy of 91.09 %, while that for EEG signals has an average baseline accuracy of 58.70 %. The experiments show that the performance significantly improves when the classifiers are trained to be gender-specific – that is, there is a separate classifier for male users, and a separate classifier for female users. For the vowel phoneme recognition dataset, the average accuracy increases to 94.20 % and 95.60 %, for male only users and female-only users, respectively. As for the EEG dataset, the accuracy increases to 65.33 % for male-only users and to 70.50 % for female-only users. Performance rates using recall and precision show the same trend. A further probe is done using SOM to visualize the distribution of the sub-clusters among male and female users.