Application of image processing techniques for frog call classification

Frogs have received increasing attention due to their effectiveness for indicating the environment change. Therefore, it is important to monitor and assess frogs. With the development of sensor techniques, large volumes of audio data (including frog calls) have been collected and need to be analysed. After transforming the audio data into its spectrogram representation using short-time Fourier transform, the visual inspection of this representation motivates us to use image processing techniques for analysing audio data. Applying acoustic event detection (AED) method to spectrograms, acoustic events are firstly detected from which ridges are extracted. Three feature sets, Mel-frequency cepstral coefficients (MFCCs), AED feature set and ridge feature set, are then used for frog call classification with a support vector machine classifier. Fifteen frog species widely spread in Queensland, Australia, are selected to evaluate the proposed method. The experimental results show that ridge feature set can achieve an average classification accuracy of 74.73% which outperforms the MFCCs (38.99%) and AED feature set (67.78%).

Garden of Shrinking Violets

Description of the work Garden of Shrinking Violets is a collection of six half scale garments and three illustrations, continuing the practice-led research project into design for disassembly, developed in the work Shrinking Violets (2015). All garments are constructed in laser cut modules that enable the items to be reassembled in new combinations. The project extended the materials used to include ahimsa (peace) silk, silk organza and silk twill. The pattern pieces have internal laser cut grids of 5mm circles, allowing the textiles to be layered, threaded and knotted to achieve rich embellished surfaces that play with the transparencies and colour overlays of the sheer and opaque silks. Research Background Conceptually grounded in design for sustainability, the aim of the work is to develop approaches to garment construction that could allow users to engage with the garments by adding, removing and reconfiguring elements. This approach to design considers the use and end-of-life phases of the transient fashion garment through considering how the garments can be later disassembled. Research Contribution This construction process is unique in being not only a patterning device but also integral to the garment’s construction. This work sits at the intersection of technical design and craft: the laser cutting and technical approach to developing new forms of garment construction is coupled with the artisanal approach of hand-knotting, a reference to traditional quilting techniques, as a method to layer and pattern the textiles. The technique developed in Shrinking Violets was extended to experiment with different grid structures, knotting devices, and decorative fringing. The result is a proposed construction system in which the laser cut grid and knotting form a decorative patterning device, but are also integral to the garments’ construction. Research Significance Garden of Shrinking Violets was exhibited at artisan gallery’s Ivory Street window, Brisbane, January 18 – February 28 2016. The work was selected by artisan gallery exhibition curators. As part of artisan gallery’s public programming, the author participated in a panel discussion: ‘Constructive conversations: deconstruction and reconstruction in contemporary craft and design’ with jeweller Elizabeth Shaw and visual arts lecturer Courtney Pedersen, 20 February 2016. Photography used in illustrations by Jonathan Rae

Deeper and wider fully convolutional network coupled with conditional random fields for scene labeling

Deep convolutional neural networks (DCNNs) have been employed in many computer vision tasks with great success due to their robustness in feature learning. One of the advantages of DCNNs is their representation robustness to object locations, which is useful for object recognition tasks. However, this also discards spatial information, which is useful when dealing with topological information of the image (e.g. scene labeling, face recognition). In this paper, we propose a deeper and wider network architecture to tackle the scene labeling task. The depth is achieved by incorporating predictions from multiple early layers of the DCNN. The width is achieved by combining multiple outputs of the network. We then further refine the parsing task by adopting graphical models (GMs) as a post-processing step to incorporate spatial and contextual information into the network. The new strategy for a deeper, wider convolutional network coupled with graphical models has shown promising results on the PASCAL-Context dataset.