inf_vb_noisy.sh

stage=$1
ckp=$2
task=$3 #"vocoder" or "se"
model_name=$4
schedule=$5 #"fast" or "full"
noisy_place=$6 # "none" or "noisy_in" or "noisy_out" or "noisy_inout"

. ./path.sh

voicebank_noisy="${voicebank}/noisy_testset_wav_16k"
voicebank_clean="${voicebank}/clean_testset_wav_16k"


if [[ ! " se vocoder " =~ " $task " ]]; then
  echo "Error: \$task must be either se or vocoder: ${task}"
  exit 1;
fi

noisy_command="--${noisy_place}"

if [[ "$noisy_place" == "none" ]]; then
    noisy_command=""
fi


if [[ "$task" == "se" ]]; then
    wav_root=${voicebank_noisy}
    spec_root=${diffwave}/spec/voicebank_Noisy_Test
    spec_type="noisy spectrum"

elif [[ "$task" == "vocoder" ]]; then
    wav_root=${voicebank_clean}
    spec_root=${diffwave}/spec/voicebank_Clean_Test
    spec_type="clean Mel-spectrum"
fi


if [ ${stage} -le 1 ]; then
    echo "stage 1 : preparing testing data"

    spec_path=${spec_root}
    wave_path=${wav_root}
    echo "create ${spec_type} from ${wave_path} to ${spec_path}"
    rm -r ${spec_path} 2>/dev/null
    mkdir -p ${spec_path}
    python src/diffwave/preprocess.py ${wave_path} ${spec_path} --${task} --test --voicebank
fi

if [ ${stage} -le 2 ]; then
    echo "stage 2 : inference model"
    target_wav_root=${voicebank_clean}

    test_spec_list=${spec_root}
    
    enhanced_path=${diffwave}/Enhanced/${model_name}/model${ckp}_${schedule}_${noisy_place}/test
    rm -r ${enhanced_path} 2>/dev/null
    mkdir -p ${enhanced_path} 
    echo "inference enhanced wav file from ${spec_root} to ${enhanced_path}"
    
    python src/diffwave/inference_noisy.py  ${diffwave}/${model_name}/weights-${ckp}.pt ${test_spec_list} ${voicebank_noisy} -o ${enhanced_path}  --${task} --voicebank --${schedule} ${noisy_command}
fi
if [ ${stage} -le 3 ]; then
    echo "stage 3 : scoring"
    
    score_file=${diffwave}/Enhanced/${model_name}/scores_test.csv
    clean_wav=${voicebank_clean}
    enhanced_result=${diffwave}/Enhanced/${model_name}/model${ckp}_${schedule}_${noisy_place}/test
    echo "save the score to ${score_file}"
    cd pesq/speech-metrics/
    python main.py ${clean_wav} ${enhanced_result} ${score_file} model${ckp}_${schedule}_${noisy_place}
    cd -
fi