Diffusion Transformer
PulseAugur coverage of Diffusion Transformer — every cluster mentioning Diffusion Transformer across labs, papers, and developer communities, ranked by signal.
- instance of Krea.2 90%
- used by Krea.2 90%
- used by Krea 2 Turbo 90%
- developed by Krea 2 Turbo 90%
- developed by Wan-Animate-2 90%
- instance of Krea 2 Turbo 90%
- used by Wan-Animate-2 90%
- used by text-to-video generation 80%
- used by Influence Flower 70%
- instance of Influence Flower 70%
- used by video diffusion transformer 70%
- instance of video diffusion transformer 70%
22 day(s) with sentiment data
-
PoseRieFlow framework generates detailed 3D human poses from text
Researchers have developed PoseRieFlow, a novel framework for generating full-body 3D human poses from text descriptions. Unlike previous methods that often overlook hand articulation or model rotations in Euclidean spa…
-
UniRGBA framework unifies RGB and alpha channels in video diffusion models
Researchers have developed UniRGBA, a novel framework that unifies RGB and alpha channel generation in pretrained video diffusion models. Unlike previous methods that required specialized architectures or matting-style …
-
Diffusion Transformer models may natively parse multiple LoRAs by 2027
A Reddit user is speculating about the future capabilities of Diffusion Transformer (DiT) models, specifically questioning if by 2027, these models will be able to natively interpret complex text prompts involving multi…
-
Diffusion Transformer interpretation challenged by new research
A new paper published on arXiv challenges the conventional understanding of diffusion transformer backbones, suggesting that their intermediate activations are not primarily indicative of changing signal granularity. In…
-
ArtCrafter generates articulated 3D objects from images
Researchers have developed ArtCrafter, a novel feed-forward generative model capable of creating articulated 3D objects from just two images. This model jointly generates part meshes for both rest and open articulation …
-
DualCast framework enhances digital human e-commerce video generation
Researchers have introduced DualCast, a novel framework designed to improve the generation of digital human e-commerce videos. This method explicitly addresses the distinct roles of product appearance and human identity…
-
Kling report details multimodal understanding for video generation
A technical report details the development of Kling, a multimodal understanding stack designed to enhance controllable video generation. The report outlines Kling's approach to linking caption density, perceptual text e…
-
New embodied AI models advance robot perception and video generation · 5 sources tracked
Researchers have introduced several new models for embodied AI and video generation, focusing on improving how robots perceive and interact with their environment. AuroraWorld aims to synthesize egocentric videos from e…
-
DiTGaze system enhances few-shot gaze estimation using diffusion models
Researchers have developed DiTGaze, a novel system for few-shot gaze estimation that utilizes a two-stream 3D Gaussian Splatting backbone and a geometry-conditioned Diffusion Transformer. This approach reformulates the …
-
Coarse3D model generates controllable and faithful sparse 3D structures from images
Researchers have introduced Coarse3D, a novel model designed to generate controllable and faithful sparse 3D structures from single images. This approach addresses limitations in existing coarse-stage generators, which …
-
SeMo3D framework unifies 3D generation, editing, and completion
Researchers have introduced SeMo3D, a novel framework that unifies 3D generation, editing, and completion tasks. By employing a "Selective Modeling" paradigm, SeMo3D focuses on synthesizing designated regions of a 3D as…
-
Diffusion Transformer framework enhances proton therapy dose denoising
Researchers have developed a diffusion-transformer framework to denoise low-statistics Monte Carlo simulations used in proton therapy. This method significantly reduces computation time while maintaining accuracy in dos…
-
OpenMOSS releases MOSS-Image text-to-image model
OpenMOSS has released MOSS-Image, a 5-billion parameter text-to-image model. This model utilizes a single-stream diffusion transformer architecture. It was made available on October 16, 2026, under the Apache 2.0 license.
-
L3G-World enables real-time infinite world generation with latent 3D Gaussian memory
Researchers have developed L3G-World, a novel system for real-time, infinite world generation that utilizes a compact latent 3D Gaussian memory. This approach addresses the challenge of maintaining spatial consistency i…
-
New XTrack Framework Leverages Video Diffusion Transformers for Object Correspondence
Researchers have developed XTrack, a novel framework that repurposes a pretrained video Diffusion Transformer (DiT) to solve cross-view temporal object correspondence. This method aims to maintain object identity across…
-
StableHand framework enhances dual-hand motion estimation from egocentric video
Researchers have developed StableHand, a novel framework for estimating dual-hand motion in world space from egocentric video. This method addresses challenges like hands leaving the camera view and severe occlusions by…
-
New ToDoPro method accelerates reference-based image generation
Researchers have developed ToDoPro, a novel method to accelerate reference-based image generation models. This technique reduces computational complexity by performing a few denoising steps on downsampled reference imag…
-
Open-source AI tools focus on infrastructure, agents, and integration
The open-source landscape for AI applications is shifting from developing new models to building the surrounding infrastructure and tools. Projects like Oumi offer a comprehensive toolchain for data preparation, fine-tu…
-
PixGR framework enhances 3D reconstruction with pixel-space generative modeling
Researchers have introduced PixGR, a novel framework that enhances 3D Gaussian reconstruction by incorporating pixel-space generative modeling. This approach models scene uncertainty and allows for efficient, end-to-end…
-
AI research explores new perception networks, visual simulation, and cybersecurity challenges
Researchers have developed TriDifNet, a novel diffusion-based network designed for end-to-end RGB-T perception, which jointly optimizes registration, fusion, and detection for misaligned infrared and visible images. Sep…