Image Processing

Deep learning has revolutionized the field of robotics. To deal with the lack of annotated training samples for learning deep models in robotics, Sim-to-Real transfer has been invented and widely used. However, such deep models trained in simulation environment typically do not transfer very well to the real world due to the challenging problem of “reality gap”. In response, this letter presents a conceptually new Digital Twin (DT)-CycleGAN framework by integrating the advantages of both DT methodology and the CycleGAN model so that the reality gap can be effectively bridged.

The ability to perceive human facial emotions is an essential feature of various multi-modal applications, especially in the intelligent human-computer interaction (HCI) area. In recent decades, considerable efforts have been put into researching automatic facial emotion recognition (FER). However, most of the existing FER methods only focus on either basic emotions such as the seven/eight categories (e.g., happiness, anger and surprise) or abstract dimensions (valence, arousal, etc.), while neglecting the fruitful nature of emotion statements.

Pelagic fish such as mackerel are a source of protein in Indonesia. However, there is no decapterus macarellus as an open dataset for image processing using various classification algorithms. Where its use includes the sensor-assisted sorting process in checking fresh fish and rotten fish. For this reason, this study aims to provide a classification model for pelagic fish and their primary datasets which is available for free on the IEEE data port. Artificial intelligence is used in the process of guided classification with the help of ground truth for the preparation of fish classes.

Fabdepth HMI is designed for hand gesture detection for Human Machine Interaction. It contains total of 8 gestures performed by 150 different individuals. These individuals range from toddlers to senior citizens which adds diversity in this dataset. These gestures are available in 3 different formats namely resized, foreground=-background separated and depth estimated images. Additional aspect is added in terms of video format of 150 samples. Researchers may choose their combination of data modalities based on their application.