| { |
| "field_id": "F020", |
| "field": "audio signal processing", |
| "directions": [ |
| { |
| "direction_id": "G0", |
| "name": "Sound-source separation and extraction", |
| "scope": "Recover constituent sounds or the audio of a specified target from a mixed recording. This fully encompasses conventional separation, explicit target cues, audiovisual source association, and separation based on direction or spatial region; these branches need not use the same cues or input modalities." |
| }, |
| { |
| "direction_id": "G1", |
| "name": "Audio enhancement and restoration", |
| "scope": "Use an existing recording or a previously estimated signal to suppress degradation or reconstruct missing content, producing more faithful and usable audio." |
| }, |
| { |
| "direction_id": "G2", |
| "name": "Sound generation and representation-based rendering", |
| "scope": "Generate audio from sound-producing parameters, acoustic features, or learned representations. This encompasses both editable sound design and faithful waveform reconstruction; an individual study need not pursue both controllability and reconstruction fidelity." |
| }, |
| { |
| "direction_id": "G3", |
| "name": "Spatial acoustic scene characterization", |
| "scope": "Characterize relationships between sound, spatial location, and the propagation environment. This includes both inferring sound-source positions and trajectories and reconstructing or rendering acoustic responses that vary with direction and geometry; neither output is a prerequisite for the other." |
| }, |
| { |
| "direction_id": "G4", |
| "name": "Acoustic recognition and classification", |
| "scope": "Determine the categories or identities of sound events, scenes, speakers, or source conditions from recorded audio, rather than primarily recovering their waveforms or spatial positions." |
| } |
| ] |
| } |
|
|
