Cell segmentation is a challenging task due to the imaging modality employed, cell’s deformable construct and imaging settings. The use of Deep Neural Networks (DNN) has facilitated excellent results in segmentation when compared to traditional image segmentation techniques as it has shown significant improvement in several metrics including accuracy, efficiency, precision etc. From the end user perspective, a good trained model is one that has a high accuracy, is portable, and has a short execution time. The size of the trained model is essential in the age of tiny machine learning (TinyML) and several techniques are developed. Among the techniques that enable a good performance while operating with small size of the input data and small size of output data we can list data augmentation, pruning and quantization. Here we report results of cell segmentation while using pruning and segmentation.