DECtalk Robot Voice Dataset: Source Audio and Mixed Audio with EnvSDD Backgrounds

This dataset contains DECtalk-synthesized robot voice audio (1000 files) and mixed audio combining robot voice with EnvSDD real/fake environmental backgrounds at 2s and 4s durations. Three volume balance conditions are provided: balanced (60dB/60dB), voice_dominant (80dB/60dB), and bg_dominant (60dB/80dB). All audio is 16000 Hz mono WAV. Includes synthesis code for reproducibility. Contents: - robot_voice_source.zip: 1000 DECtalk robot voice files (avg 7.75s) - robot_voice_real_bg_2s.zip: 3000 mixed files (robot voice + real EnvSDD bg, 2s) - robot_voice_fake_bg_2s.zip: 3000 mixed files (robot voice + fake EnvSDD bg, 2s) - robot_voice_real_bg_4s.zip: 3000 mixed files (robot voice + real EnvSDD bg, 4s) - robot_voice_fake_bg_4s.zip: 3000 mixed files (robot voice + fake EnvSDD bg, 4s) - robot_voice_code.zip: Python synthesis scripts