<p>Thanks for your answer :)</p>

<blockquote>
<p>It's worth noting that the GMM code you pasted where you iterate over each image and call gmm-&gt;Train(images[i], 1, true) is actually quite different than what hmm-&gt;Train(images) will do.</p>
</blockquote>

<p>Yes, I figured that out eventually and flattened all my sequence into one. </p>

<blockquote>
<p>Another thing worth noting is that, in general, training a GMM-HMM without labeled data is a long process that may not give very good results. It's much better to do labeled training if possible.</p>
</blockquote>

<p>Does that change anything when there is only one state ? </p>

<p style="font-size:small;-webkit-text-size-adjust:none;color:#666;">&mdash;<br>Reply to this email directly or <a href="https://github.com/mlpack/mlpack/issues/550#issuecomment-196238987">view it on GitHub</a>.<img alt="" height="1" src="https://github.com/notifications/beacon/AJ4bFG0KG95DCkY82AZrZT_kiw6FF6D0ks5ptTTHgaJpZM4HpaV-.gif" width="1" /></p>
<div itemscope itemtype="http://schema.org/EmailMessage">
<div itemprop="action" itemscope itemtype="http://schema.org/ViewAction">
  <link itemprop="url" href="https://github.com/mlpack/mlpack/issues/550#issuecomment-196238987"></link>
  <meta itemprop="name" content="View Issue"></meta>
</div>
<meta itemprop="description" content="View this Issue on GitHub"></meta>
</div>