Speaker identification systems in a real-world scenario are tasked to identify a speaker amongst a set of enrolled speakers given just a few samples for each enrolled speaker. This paper demonstrates the effectiveness of meta-learning and relation networks for this use case. We propose improved relation networks for speaker verification and few-shot (unseen) speaker identification. The use of relation networks facilitates joint training of the frontend speaker encoder and the backend model. Inspired by the use of prototypical networks in speaker verification and to increase the discriminability of the speaker embeddings, we train the model to classify samples in the current episode amongst all speakers present in the training set. Furthermore, we propose a new training regime for faster model convergence by extracting more information from a given metalearning episode with negligible extra computation. We evaluate the proposed techniques on VoxCeleb, SITW and VCTK datasets on the tasks of speaker verification and unseen speaker identification. The proposed approach outperforms the existing approaches consistently on both tasks.

No abstract

Speaker identification systems are deployed in diverse environments, often different from the lab conditions on which they are trained and tested. In this paper, first, we show the problem of generalization using fixed thresholds computed using the equal error rate metric. Secondly, we introduce a novel and generalizable speaker-specific thresholding technique for robust imposter identification in unseen speaker identification. We propose a speaker-specific adaptive threshold, which can be computed using the enrollment audio samples, for identifying imposters in unseen speaker identification. Furthermore, we show the efficacy of the proposed technique on VoxCeleb1, VCTK and the FFSVC 2022 datasets, beating the baseline fixed thresholding by up to 25%. Finally, we exhibit that the proposed algorithm is also generalizable, demonstrating its performance on ResNet50, ECAPA-TDNN and RawNet3 speaker encoders.

In the broader picture, Interpretation of Statutes donnes a rather infinitesimal scope to its applications in the practical judicial setup of the nation. Be it the Constitution or the plethora of laws that are passed by the parliament to run the nation and govern its various stratum, interpretation plays a dominant role in getting the same to percolate down to the grass-root level. Benjamin N Cardozo has very famously theorized a lot of what the scope of interpretation of statutes entails and has even streamlined major theories to deduce the process of derivation used by judges to draw out conclusions in various cases. These postulations, although not expressly, find themselves being effectuated via various judges and the same can be deduced upon closer inspection of the wordings of the judgment a judge chooses to articulate while delivering the same. These processes of reasoning by judges vary in their applications and lead to consequences that have been intended by the legislature or rather furthered by the principles of justice, equity and good conscience, the three pillars of a just society. Through the course of this paper, we analyse two major judgments, Sowmithri Vishnu v. Union of India and Joseph Shine v Union of India. Both cases were instituted challenging the constitutionality of section 497 of the Indian Penal Code, that criminalise the act of adultery. The former case, adjudged by Justice YV Chandrachud upheld its constitutional validity whereas the later one adjudged by Justice DY Chandrachud decriminalised the section. Adjudging the interpretative paths taken by the two judges in light of Cardozo’s understanding of Interpretation provides unique insight into the workings of the judicial minds and how time influence the scope of the same. The variance in judgments delivered by father and son presents as an important ground to understand the role of interpretation of statutes and its applications in the world while also helps map out a certain evolutionary process that may be attributed to the scope of interpretation in light of various societal factors that also affect it.