Pradip Sasmal

Avinash Amballa, Advaith P., Pradip Sasmal et al.

Training self-driving cars is often challenging since they require a vast amount of labeled data in multiple real-world contexts, which is computationally and memory intensive. Researchers often resort to driving simulators to train the agent and transfer the knowledge to a real-world setting. Since simulators lack realistic behavior, these methods are quite inefficient. To address this issue, we introduce a framework (perception, planning, and control) in a real-world driving environment that transfers the real-world environments into gaming environments by setting up a reliable Markov Decision Process (MDP). We propose variations of existing Reinforcement Learning (RL) algorithms in a multi-agent setting to learn and execute the discrete control in real-world environments. Experiments show that the multi-agent setting outperforms the single-agent setting in all the scenarios. We also propose reliable initialization, data augmentation, and training techniques that enable the agents to learn and generalize to navigate in a real-world environment with minimal input video data, and with minimal training. Additionally, to show the efficacy of our proposed algorithm, we deploy our method in the virtual driving environment TORCS.

Pradip Sasmal, Phanindra Jampana, C. S. Sastry

The exact recovery property of Basis pursuit (BP) and Orthogonal Matching Pursuit (OMP) has a relation with the coherence of the underlying frame. A frame with low coherence provides better guarantees for exact recovery. In particular, Incoherent Unit Norm Tight Frames (IUNTFs) play a significant role in sparse representations. IUNTFs with special structure, in particular those given by a union of several orthonormal bases, are known to satisfy better theoretical guarantees for recovering sparse signals. In the present work, we propose to construct structured IUNTFs consisting of large number of orthonormal bases. For a given $r, k, m$ with $k$ being less than or equal to the smallest prime power factor of $m$ and $r<k,$ we construct a CS matrix of size $mk \times (mk\times m^{r})$ with coherence at most $\frac{r}{k},$ which consists of $m^{r}$ number of orthonormal bases and with density $\frac{1}{m}$. We also present numerical results of recovery performance of union of orthonormal bases as against their Gaussian counterparts.

Laura Rebollo-Neira, Miroslav Rozloznik, Pradip Sasmal

The convergence and numerical analysis of a low memory implementation of the Orthogonal Matching Pursuit greedy strategy, which is termed Self Projected Matching Pursuit, is presented. This approach renders an iterative way of solving the least squares problem with much less storage requirement than direct linear algebra techniques. Hence, it appropriate for solving large linear systems. The analysis highlights its suitability within the class of well posed problems.