AI That Fixes Your Eye Contact in Photos—No Labels Needed

Kari Jaaskelainen

12 Nov 2025 — 1 min read

Look me in the eye, camera.

Ever snapped a great photo except your eyes aren’t on the camera? This research introduces GazeCorrection, an AI that subtly redirects your gaze by re-synthesizing just the eye region—keeping your identity and expression intact.

How it works: a self-supervised generative model treats gaze editing like smart inpainting. Instead of needing labeled head poses or eye angles, it learns from unlabeled face images to fill in the “missing” eye area with a corrected, forward-looking gaze. A self-guided pretraining step helps the model capture angle-invariant features, so results look natural “in the wild” (photos scraped from the web).

Label-free training data
Identity-preserving edits
Compelling results on diverse, real-world images

Use cases span video calls, group photos, accessibility, and content creation. As with any face-editing tech, ethical, consent-based use is essential.

Paper: arxiv.org/abs/1906.00805 • Code: github.com/zhangqianhui/GazeCorrection

Paper: http://arxiv.org/abs/1906.00805v1

Register: https://www.AiFeta.com

AI ComputerVision GAN GazeCorrection EyeContact Inpainting SelfSupervised Research OpenSource Ethics

Automating GDPR Compliance: A Roadmap for Companies and Law Firms

GDPR compliance is more than checkboxes. A new roadmap from the Privatech project shows how automation and machine learning can help companies and law firms assess—and even generate—privacy compliance. * Shift the focus to data processors’ real workflows: drafting policies, mapping data uses, documenting decisions. * Break compliance into machine-ready

FPGAs for Faster, Leaner Deep Learning: A Review of CNN Accelerators

Deep learning drives image search, robots, and medical scans. Most systems lean on CPUs and GPUs. This review asks: what if we run convolutional neural networks (CNNs) on FPGAs—reconfigurable chips you can tailor to the model? * Why FPGAs: custom dataflows, low latency, and strong energy efficiency—great for cameras,

Dynamic-K: Recommendations That Know When to Stop

Most apps show a fixed number of “top” items—say 10 movies or 20 products—assuming there are always enough good options. But that’s not always true: sometimes there are few relevant items, or some users are extra picky. The result? Filler recommendations. Dynamic-K flips the script. Instead of

Teaching chatbots to stop contradicting themselves (DECODE)

Teaching chatbots to stop contradicting themselves Ever had a bot say one thing, then the opposite a few turns later? This study introduces DECODE—a new task and dataset for spotting contradictions in everyday conversations, drawn from both human-human and human-bot chats. * New data beats existing natural language inference (NLI)

Look me in the eye, camera.

Read more

Automating GDPR Compliance: A Roadmap for Companies and Law Firms

FPGAs for Faster, Leaner Deep Learning: A Review of CNN Accelerators

Dynamic-K: Recommendations That Know When to Stop

Teaching chatbots to stop contradicting themselves (DECODE)