Timeshrinking

From SpeechWiki

Revision as of 01:29, 2 October 2009 (view source) Arthur (Talk | contribs) ← Older edit		Revision as of 15:23, 2 October 2009 (view source) Arthur (Talk | contribs) Newer edit →
Line 32:		Line 32:
	| 1 || 1 || [{{FisherPath}}/exp/timeshrink/test/.../LATEST.log TR] || [{{FisherPath}}/exp/timeshrink/test/triphoneSingleGausian/unit.tri.timeshrink.1.mlp.onSingleGaussian/LATEST.log 72.8% ] || PLP+MLP tandem		| 1 || 1 || [{{FisherPath}}/exp/timeshrink/test/.../LATEST.log TR] || [{{FisherPath}}/exp/timeshrink/test/triphoneSingleGausian/unit.tri.timeshrink.1.mlp.onSingleGaussian/LATEST.log 72.8% ] || PLP+MLP tandem
	|-		|-
-	|}	+	| .9 || .9 || [{{FisherPath}}/exp/timeshrink/test/.../LATEST.log TR] || [{{FisherPath}}/exp/timeshrink/test/triphoneSingleGausian/unit.tri.timeshrink.point9.mlp.onSingleGaussian/LATEST.log 73.1% ] || PLP+MLP tandem
		+	|-|}
	I have to check for bugs.  It could be that the threshold is too low or it could be something else too.  We should probably rerun baseline too, just to make sure I didn't optimize it unfairly.		I have to check for bugs.  It could be that the threshold is too low or it could be something else too.  We should probably rerun baseline too, just to make sure I didn't optimize it unfairly.

---

Revision as of 15:23, 2 October 2009

Fisher experiments

number of frames dropped on Fisher corpus

<math>\tau</math>	frames dropped
1	0%
.9	~5%
.6	~35%

I have to check for bugs. It could be that the threshold is too low or it could be something else too. We should probably rerun baseline too, just to make sure I didn't optimize it unfairly.

Things to try

• Test svitchboard with fisher-trained model to see if we still get good results
• Train and test on plp+mlp, like svitchboard timeshrinking was done.
• Do baseline train+test to see if something changed in going from baseline to timeshrink structure files.

Future Directions

• Can be viewed as a two-mode special case of best-first viterbi search. So make a real best-first lattice search. Mark mentioned some attempts in the 80'ies to do this.

Timeshrinking results on fisher

train <math>\tau</math>	test <math>\tau</math>	dev 2000 utt WER	dev 2000 utt on triphone single-gaussian model WER	comments
1	1	51.6%		old baseline
1	1	53.7%	78.2%	baseline rerun exactly as timeshrinking to really make sure it's not getting an unfair advantage
.6	.6	69.3%
.9	.9	56.3%	80.4%
1	.9	53.9%
.9	.9	57.2	80.7%	using the non-timeshrinking str file for test
.9	1	54.6
1	1	TR	72.8%	PLP+MLP tandem
.9	.9	TR	73.1%	PLP+MLP tandem