In Part 1 I found that my Malayalam speech model’s test set leaked into its training data. In Part 2 I re-tested the model on clean data only, and promised a better v2. This post reports two things. First, v2 is trained, and on the same clean test set it cuts word errors by 12
Part 3: Retraining My Malayalam ASR Model, and a Bug in My Own Evaluation Read More »
Correction (11 October 2026): The v1 numbers in this post came from an evaluation notebook that capped transcripts at 225 tokens. Malayalam needs about 25 Whisper tokens per second of speech, so clips longer than about 9 seconds were cut off and the missing words counted as errors. With the cap fixed, v1 scores 43.5%
I Re-Tested My Malayalam ASR Model on Clean Data — Part 2 Read More »
Update (October 2026): This is Part 1 of a three-part series. Part 2 re-tests v1 on the clean test set built here. Part 3 reports the finished v2 (43.5% → 31.2% WER) and a bug in my own evaluation. Three things in this post have changed since I wrote it: v2 is trained. It used
I Checked My Own ASR Dataset for Leakage — Here’s What I Found Read More »
Introduction TinyML is the field which involves deployment of machine learning models into resource constrained devices such as micro-controllers. Such devices called edge devices often have few Kilobytes of RAM and flash memory but consumes power in milli-watts range. This feature makes the technology an ideal choice for remote sensing applications, weather stations, tiny gadgets
Approximating Sine Function in TinyML Read More »
User Interface is key for humans to interact with computers. The traditional way is to use keyboard and mouse to browse websites, use apps, or play video games. When it comes to specific applications such as infographics display in a mall or airport often the amount of content will be limitted and simple. In such
User Interface Control Using Hand Gesture Recognition Read More »
The use of computational tools for data preprocessing and analysis is daily activity for any bioinformatician. Having a interdisciplinary field, a plethora of tools exist in the domain, written by developers from various fields and languages. Many times there can be multiple candidate tools exists to accomplish any given task. Often it is the decision
Writing Python Wrapper for Bioinformatics Tools Read More »
The field of information theory defines entropy as a measure of average information a random variable takes. As we know the random variable is one whose possible outcome states are not deterministic. We cannot predict its behavior exactly in an analytical form. In such cases, we can only look at the patterns of its outcome
Entropy Simplified Read More »
This article is a tutorial on the algorithm called active Learning with uncertainty Sampling. Introduction Availability of mass quantities of digital data and feasible computing power brought to the creation of learning algorithms. These learning algorithms have been benchmarked to perform specialized tasks such as classification, object detection, image segmentation, etc. The key assumption here
Active Learning with Uncertainty Sampling from Scratch Read More »
In one of our earlier posts, we have seen how we can visually relate the parts of the one-dimensional Gaussian distribution equation. In this post, we will follow the same strategy to understand the terms that comes up with a Multivariable Gaussian distribution. We will focus on the Bivariate Gaussian distribution as distributions of higher-order
Hacking the Bivariate Gaussian Distribution Read More »
This article explains the minmax scaling operation using visual examples. Normalization of vectors, an array of values, signals is often used as a preprocessing step before many algorithms. For example, in machine learning, some types of algorithms are prone to different inherent scales of features. In such situations normalization is done to give the same
Visualizing MinMax Scaling Read More »