This thesis develops a Transformer model based on Whisper, which extracts melodies and chords from music audio and records them into ABC notation. A comprehensive data processing workflow is customized for ABC notation, including data cleansing, formatting, and conversion, and a mutation mechanism is implemented to increase the diversity and quality of training data. This thesis innovatively introduces the "Orpheus' Score", a custom notation system that converts music information into tokens, designs a custom vocabulary library, and trains a corresponding custom tokenizer. Experiments show that compared to traditional algorithms, the model has significantly improved accuracy and performance. While providing a convenient audio-to-score tool for music enthusiasts, this work also provides new ideas and tools for research in music information processing.

On May 14, the National Oceanic and Atmospheric Administration (NOAA) ship Okeanos Explorer will depart from Port Canaveral in Florida on a two-week expedition led by NOAA Ocean Exploration, featuring the technology demonstration of an autonomous underwater vehicle. Called Orpheus, this new class of submersible robot will showcase a system that will help it find its way and identify interesting scientific features on the seafloor. Terrain-relative navigation was instrumental in helping NASA's Mars 2020 Perseverance Mars rover make its precision touch down on the Red Planet on Feb. 18. The system allowed the descending robot to visually map the Martian landscape, identify hazards, and then choose a safe place to land without human assistance. In a similar way, the agency's Ingenuity Mars Helicopter uses a vision-based navigation system to track surface features on the ground during flight in order to estimate its movements across the Martian surface.

Optimising deep learning inference across edge devices and optimisation targets such as inference time, memory footprint and power consumption is a key challenge due to the ubiquity of neural networks. Today, production deep learning frameworks provide useful abstractions to aid machine learning engineers and systems researchers. However, in exchange they can suffer from compatibility challenges (especially on constrained platforms), inaccessible code complexity, or design choices that otherwise limit research from a systems perspective. This paper presents Orpheus, a new deep learning framework for easy prototyping, deployment and evaluation of inference optimisations. Orpheus features a small codebase, minimal dependencies, and a simple process for integrating other third party systems. We present some preliminary evaluation results.

Training on only perfect Standard English corpora predisposes pre-trained neural networks to discriminate against minorities from non-standard linguistic backgrounds (e.g., African American Vernacular English, Colloquial Singapore English, etc.). We perturb the inflectional morphology of words to craft plausible and semantically similar adversarial examples that expose these biases in popular NLP models, e.g., BERT and Transformer, and show that adversarially fine-tuning them for a single epoch significantly improves robustness without sacrificing performance on clean data.

Whenever I think about robots, I can't help but think about Terminator. You know, the one that wants to rid the earth of humankind? Well, this mechanical companion from ThinkGeek flips the script on the maniacal machine cliché. His name is Orpheus, and he is a DIY wind-up robot that doubles as a wind-up music box. The kit comes with 220 laser-cut wood pieces and you're supposed to build the entire thing yourself.