Piyush Thakur

Table of Contents Grounded SAM 2: From Open-Set Detection to Segmentation and Tracking Why Segmentation Matters (Beyond Bounding Boxes) Introducing Grounded SAM 2 Where SAM Fits in the Pipeline Why SAM 2 (and not SAM) How Grounded SAM 2 Works…

Table of Contents Grounding DINO: Open Vocabulary Object Detection on Videos Why Language Makes Open-Set Detection Possible GLIP: Grounded Language-Image Pre-Training The DINO Detector (Closed-Set DETR) Grounding DINO Architecture Feature Enhancer (Neck Fusion) and Cross-Attention: The Teacher’s Guidance Language-Guided Query…

Table of Contents Build a VLC Playlist Generator with SmolVLM for Video Highlight Tagging Configuring Your Development Environment Setup and Imports Helper Functions Main VideoHighlightDetector Class Creating the XSPF Playlist Gradio Interface Logic Launch the Gradio Application Output Summary Citation…

Table of Contents Running SmolVLM Locally in Your Browser with Transformers.js Introduction SmolVLM: A Small But Capable Vision-Language Model Transformers.js: Hugging Face Runtime for the Web Why This Pair Is Powerful (and Practical Caveats) Configuring Your Development Environment What Are…

Table of Contents Synthetic Data Generation Using the VLM-as-Judge Method Configuring Your Development Environment Set Up and Imports Downloading Images Locally Using Qwen as VLM-as-Judge Converting the JSON File to the Hugging Face Dataset Format Inspecting the Dataset Pushing the…

Table of Contents Synthetic Data Generation Using the BLIP and PaliGemma Models Why VLM-as-Judge and Synthetic VQA Configuring Your Development Environment Set Up and Imports Download Images Locally Inference with the Salesforce BLIP Model Convert JSON File to the Hugging…

Table of Contents Generating Video Highlights Using the SmolVLM2 Model Configuring Your Development Environment Setup and Imports Setup Logger Get Video Duration in Seconds Load Model and Processor Analyze Video Content Determine Highlights Process Video Segment Concatenating Video Scenes into…

Table of Contents SmolVLM to SmolVLM2: Compact Models for Multi-Image VQA SmolVLM 1: A Compact Yet Capable Vision-Language Model What Is SmolVLM? Why SmolVLM? The Three Variants of SmolVLM Architecture Overview Vision Encoder: SigLIP Variants Pixel Shuffle (Space-to-Depth) for Image…

Table of Contents AI for Healthcare: Fine-Tuning Google’s PaliGemma 2 for Brain Tumor Detection Configuring Your Development Environment Setup and Imports Load the Brain Tumor Dataset Format Dataset to PaliGemma Format Display Train Image and Label COCO Format BBox to…