Category: Computer Vision

Computer Vision in AI encompasses various tasks including classical computer vision, deep learning based image classification, image classification, object detection etc.

Self-Hosting GLM-OCR using vLLM – Document Layout and OCR

In this article, we explore self-hosting GLM-OCR with document layout parsing via vLLM. We also compare the results with the non-document layout pipeline where only OCR happens. ...

SAM 3 UI – Image, Video, and Multi-Object Inference

Sovit Ranjan Rath February 23, 2026 0 Comment

In this article, we create a simple SAM 3 Gradio UI for image and video segmentation. SAM 3 UI supports segmenting objects belonging to the different categories while using less than 10GB VRAM. ...

SAM 3 Inference and Paper Explanation

Sovit Ranjan Rath February 9, 2026 2 Comments

In this article we cover the SAM3 model. We discuss the SAM3 paper briefly including the motivation, the architecture, and the data engine. Next, we move on to image and video inference using SAM3. ...

Hunyuan3D 2.0 – Explanation and Runpod Docker Image

Sovit Ranjan Rath February 2, 2026 0 Comment

In this article, we cover the explanation of the Hunyuan3D 2.0 technical report and create a Runpod Docker Image for the same for smoother execution of image-to-3D workflows. ...

Image-to-3D: Incremental Optimizations for VRAM, Multi-Mesh Output, and UI Improvements

Sovit Ranjan Rath January 26, 2026 0 Comment

In this article we carry out optimizations for the image-to-3D pipeline in terms of VRAM usage, multi-object generation from prompts, and improved UI. The pipeline uses Qwen3-VL, BiRefNet, and Hunyuan3D models. ...