Abstract: The improvement of automatic transcription algorithms has led to the accessibility of high-quality symbolic music data aligned with audio, especially for piano music. However, the transcribed MIDI files need extra annotation, such as beat, chord, and structure, to utilize the dataset in other MIR tasks, such as automatic music composition, melody harmonization, and harmonic analysis. Music annotations, especially chord annotations, need high-level music domain knowledge. Although there are a handful of algorithms for chord annotation, the prediction accuracy is limited. Therefore, it needs humans to correct the annotation eventually. We suggest a web-based semi-automated chord annotation tool for piano music to facilitate this validation process conveniently. We integrated the previous piano music transcription, alignment, quantization, and chord annotation research into one system. Given piano audio input, users can modify the predicted chord annotation of converted MIDI. The demo page is available in our website.