Diffusion Models Beat GANs on Image Synthesis · HackerPapers discussions

Diffusion Models Beat GANs on Image Synthesis — mentioned by godelski

Fri, 27 Jun 2025 23:48:00 GMT

As far as I'm aware, this is the largest Normalizing Flow that exists, and I think they undermined their work by not mentioning this... Their ImageNet model (4_1024_8_8_0.05[0]) is ~820M while AFHQ is ~472M. Prior to that there is DenseFlow[1] and MaCow[2], which are both <200M parameters. For more comparison, that makes DenseFlow and MaCow smaller than iDDPM[3] (270M params) and ADM[4] (553M for 256 unconditional). And now, it isn't uncommon for modern diffusion models to have several billion parameters![5] (from this we get some numbers on ImageNet-256, which allows a direct comparison…

Paper and discussion history. A mention is not an endorsement.

Diffusion Models Beat GANs on Image Synthesis — mentioned by godelski

Tue, 06 Aug 2024 08:26:48 GMT

> The rule in ML There's definitely attempts to revive things (in the general sense, not just GANs), but most successes appear to come from large labs dedicating equal computational resources to the older models and often by changing names. This can make things more confusing and make things appear to be changing faster, but once you can see this, you'll have an easier time keeping up (so being new, watch out for this). I'll give some examples that are easier to read[0-2] (i.e. don't need expert knowledge to understand the nuances). As an insider (ML researcher), my complaint isn't so much…

Paper and discussion history. A mention is not an endorsement.

Diffusion Models Beat GANs on Image Synthesis — mentioned by reasonableklout

Sun, 26 May 2024 21:33:14 GMT

The original Dickstein 2015 paper [1] formulated diffusion as maximizing (a lower bound of) log-likelihood of generating the distribution, so there was some theory. But my understanding is that the breakthrough was empirical results from Ho [2] and Nichol [3] showing diffusion could produce not only high-quality samples but better than GANs in some cases. [1] https://arxiv.org/abs/1503.03585 [2] https://arxiv.org/abs/2006.11239 [3] https://arxiv.org/abs/2105.05233

Paper and discussion history. A mention is not an endorsement.

Diffusion Models Beat GANs on Image Synthesis — mentioned by GaggiX

Wed, 30 Nov 2022 14:45:28 GMT

"It was abandoned without any evidence whatsoever that diffusion model etc is required.", diffusion models beats GAN models in quality on all datasets it was tested on, I suggest to read (again probably): "Diffusion Models Beat GANs on Image Synthesis" [1], the difference in quality is astounding, also the fact that GANs seems to work well with small one domain dataset, then they became incrementally more difficult to train with Imagenet and then it became easier again with 300M images only demonstrates how sensible these models are to hyperparameters and unstable they are to train (some…

Paper and discussion history. A mention is not an endorsement.

Diffusion Models Beat GANs on Image Synthesis — HN submission

Wed, 12 May 2021 01:43:18 GMT

Diffusion Models Beat GANs on Image Synthesis

209 points; 50 discussion replies on HN. Paper and source history. A submission is not an endorsement.