Bit Oasis

Business Registration Certificate - Platform Verification Approved

Education & Background

• MSc or PhD in Computer Science, Electronic Engineering, Mathematics, or related fields, from a reputable university.

• Strong academic or industry track record in computer vision, 3D vision, or multimodal learning.

Role Summary

Build vision-driven modules and evaluation pipelines for image-to-3D asset generation. You’ll own quality assessment for input images and generated 3D assets, and help close the data feedback loop with training and product teams.

Responsibilities

• Develop vision modules for object detection, segmentation, occlusion/pose estimation, and image quality scoring.

• Design systems that assess input-image suitability for 3D generation.

• Build evaluation pipelines for generated 3D assets (geometry, texture, consistency, usability).

• Improve data filtering, labeling, and quality metrics for large-scale image and 3D datasets.

• Collaborate with model training, 3D pipeline, and product teams to productionize research methods.

Requirements

• Strong foundation in computer vision: detection, segmentation, representation learning, multimodal models.

• Hands-on experience with modern vision architectures and visual embeddings.

• Proficient in Python and PyTorch.

• Familiarity with 3D formats (meshes, point clouds, GLB, OBJ, FBX).

• Ability to design practical evaluation metrics and debug model failures in the wild.

• Engineering best practices: reproducibility, documentation, code quality.

Nice to Have

• Experience with image-to-3D, text-to-3D, 3D reconstruction, or 3D asset QA.

• Familiarity with Blender, Open3D, trimesh, PyTorch3D, or similar tools.

• Background in gaming, e-commerce, AR/VR, digital assets, or UGC pipelines.