arXiv subjects

Subodh Kumar

Publications and source records attributed to Subodh Kumar.

3 recordsLinked to original sources

arXiv2026

Conflict-Free Cuts in Planar and 3-Degenerate Graphs with 1-Regular Conflicts

A conflict-free cut $F$ on a simple connected graph $G = (V, E)$ is defined as a set of edges $F \subseteq E$ such that $G-F$ is disconnected, and no two edges in $F$ are conflicting. The notion of conflicting edges is represented using an associated conflict graph $\widehat{G} = (\widehat{V}, \widehat{E})$ where $\widehat{V} = E$. Deciding if a given planar graph $G$, with an associated conflict graph $\widehat{G}$, has a conflict-free cut is known to be NP-complete, when $G$ has maximum degree four and $\widehat{G}$ is a line graph of $G$ [Bonsma, JGT 2009]. In this paper, we prove the following for the case when $\widehat{G}$ is 1-regular. * We completely resolve the complexity of the decision problem when $G$ is planar. Towards this end, we show that (a) there always exists a conflict-free cut when the graph is planar and 4-regular unless it is the octahedron graph and (b) the decision problem is NP-complete, even in the case when $G$ is planar with maximum degree 5. * We also show that the decision problem is NP-complete when $G$ is a 3-degenerate graph with maximum degree 5. This completely resolves the complexity status of the problem when $G$ is 3-degenerate. * We construct families of graphs with 1-regular conflict graphs that do not have a conflict-free cut. Our results answer the questions posed in [Rauch, Rautenbach and Souza, IPL 2025].

math.CO↗

arXiv2025

TSPulse: Tiny Pre-Trained Models with Disentangled Representations for Rapid Time-Series Analysis

Time-series tasks often benefit from signals expressed across multiple representation spaces (e.g., time vs. frequency) and at varying abstraction levels (e.g., local patterns vs. global semantics). However, existing pre-trained time-series models entangle these heterogeneous signals into a single large embedding, limiting transferability and direct zero-shot usability. To address this, we propose TSPulse, family of ultra-light pre-trained models (1M parameters) with disentanglement properties, specialized for various time-series diagnostic tasks. TSPulse introduces a novel pre-training framework that augments masked reconstruction with explicit disentanglement across spaces and abstractions, learning three complementary embedding views (temporal, spectral, and semantic) to effectively enable zero-shot transfer. In-addition, we introduce various lightweight post-hoc fusers that selectively attend and fuse these disentangled views based on task type, enabling simple but effective task specializations. To further improve robustness and mitigate mask-induced bias prevalent in existing approaches, we propose a simple yet effective hybrid masking strategy that enhances missing diversity during pre-training. Despite its compact size, TSPulse achieves strong and consistent gains across four TS diagnostic tasks: +20% on the TSB-AD anomaly detection leaderboard, +25% on similarity search, +50% on imputation, and +5-16% on multivariate classification, outperforming models that are 10-100X larger on over 75 datasets. TSPulse delivers state-of-the-art zero-shot performance, efficient fine-tuning, and supports GPU-free deployment. Models and source code are publicly available at https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1.

cs.LG↗

arXiv2022

On Comparison of Encoders for Attention based End to End Speech Recognition in Standalone and Rescoring Mode

The streaming automatic speech recognition (ASR) models are more popular and suitable for voice-based applications. However, non-streaming models provide better performance as they look at the entire audio context. To leverage the benefits of the non-streaming model in streaming applications like voice search, it is commonly used in second pass re-scoring mode. The candidate hypothesis generated using steaming models is re-scored using a non-streaming model. In this work, we evaluate the non-streaming attention-based end-to-end ASR models on the Flipkart voice search task in both standalone and re-scoring modes. These models are based on Listen-Attend-Spell (LAS) encoder-decoder architecture. We experiment with different encoder variations based on LSTM, Transformer, and Conformer. We compare the latency requirements of these models along with their performance. Overall we show that the Transformer model offers acceptable WER with the lowest latency requirements. We report a relative WER improvement of around 16% with the second pass LAS re-scoring with latency overhead under 5ms. We also highlight the importance of CNN front-end with Transformer architecture to achieve comparable word error rates (WER). Moreover, we observe that in the second pass re-scoring mode all the encoders provide similar benefits whereas the difference in performance is prominent in standalone text generation mode.

cs.SD↗