UTAU is a shareware vocal synthesizer ([vocal-synthesis]) that uses sampled voicebanks to produce synthesized singing. It seems to the leading free alternative to Vocaloid; it's relatively easy to produce voicebanks by sampling acapellas or kareoke songs with labeled lyrics. There are even tools out there to automatically produce voicebanks for UTAU from tagged sources, although I'm a bit fuzzy on the details. See Moresampler (possibly discontinued).
UTAU voicebanks consist of labeled samples of sound. There are different formats of them such as VCV (vowel-consonant-vowel), CVVC, and others which produce different output characteristics. This is a different approach from that used by [hts] which produce models rather than relying entirely on samples. As a result, they are much easier to create than HTS models making them more available for people looking to get involved.
See: http://thewesternguidetoutau.yolasite.com/voice-banks.php