Urdu Handwritten Characters Data Visualization and Recognition Using Distributed Stochastic Neighborhood Embedding and Deep Network

<div>Visualization of (a) pixel-based and (b) structure-based data of our dataset of handwritten Urdu characters using standard <i>t</i>-SNE took 12 minutes 31 seconds. (a) Perplexity: 30 Iterations: 1,000. (b) Perplexity: 50 Iterations: 1,000.</div>

Complexity

fig9

Figure 9

Figure 9: Urdu Handwritten Characters Data Visualization and Recognition Using Distributed Stochastic Neighborhood Embedding and Deep Network 