Claims Browser

The book enables understanding of various neural network architectures, including image classification, object detection, segmentation, generative adversarial networks, transformers, and diffusion models.

"🟣Understand the inner workings of various neural network architectures and their implementation, including image classification, object detection, segmentation, generative adversarial networks, transformers, and diffusion models."

Runway is giving away 1,000,000 credits through a HORSE competition challenge

"We're giving away 1,000,000 credits. All you have to do is beat us at HORSE."

Astroparticle physicists are using Keras to replace hand-crafted features and directly model raw spatio-temporal waveforms for detecting cosmic ray origins

"astroparticle physicists are using Keras to replace the usual hand-crafted features and directly model the raw spatio-temporal waveforms"

LeVJEPA is the first video encoder trained under LeJEPA's collapse-free objective that dispenses with both architectural asymmetries and pixel-space masked content reconstruction

"We introduce LeVJEPA, the first video encoder trained under LeJEPA's collapse-free objective, which dispenses with both."

LeVJEPA matches or surpasses V-JEPA 2 across ViT-S/B/L at 5.6 to 20.8x less pretraining compute when trained for matched epochs on identical data

"At matched epochs on identical data, LeVJEPA matches or surpasses V-JEPA 2 across ViT-S/B/L at 5.6 to 20.8x less pretraining compute"

At matched total FLOPs, LeVJEPA exceeds the strongest video baseline by 7.6 points on ImageNet-1K while remaining competitive on motion-centric benchmarks

"at matched total FLOPs it exceeds the strongest video baseline by 7.6 points on ImageNet-1K while remaining competitive on motion-centric benchmarks"

Uniform random token dropping reduces computational cost while simultaneously improving downstream accuracy in LeVJEPA

"uniform random token dropping renders this number small while simultaneously improving downstream accuracy"

The encoder can be trained with block-causal attention at no measurable accuracy cost, making temporal ordering a property of the encoder itself

"since no asymmetry between branches is required, the encoder can be trained with block-causal attention at no measurable accuracy cost: temporal ordering becomes a property of the encoder itself"

Against a compute-matched DINOv2 trained on frames of the same videos, LeVJEPA approaches the image-pretrained encoder on appearance-centric evaluation while nearly doubling its motion-centric accuracy

"Against a compute-matched DINOv2 trained on frames of the same videos, LeVJEPA approaches the image-pretrained encoder on appearance-centric evaluation while nearly doubling its motion-centric accuracy"

Large Reconstruction Models (LRMs) provide a powerful geometric scaffold that preserves relative human-object arrangement and proximity cues for 3D human-object interaction reconstruction

"Our key observation is that LRMs provide a powerful geometric scaffold that preserves relative human-object arrangement and proximity cues."

MILO achieves strong reconstruction accuracy and outperforms existing baselines across multiple benchmarks and interaction scenarios for 3D human-object interaction estimation

"MILO achieves strong reconstruction accuracy and outperforms existing baselines across multiple benchmarks and interaction scenarios."

Using LRMs significantly simplifies the 3D human-object interaction reconstruction procedure by reframing it as interpreting the LRM mesh rather than fitting models to 2D images

"This significantly simplifies the reconstruction procedure, reframing the problem as interpreting the LRM mesh: we segment it into human and object components, fit a parametric body model to the human part, and optionally align an object template to the object part"

XTTSv2's voice cloning capabilities can preserve prosodic structure independently of speaker identity, enabling it to be repurposed for speaker anonymization without retraining

"Our key insight is that XTTSv2's voice cloning capabilities preserve prosodic structure independently of speaker identity, enabling voice conversion by conditioning on a pseudo-speaker."

The proposed anonymization system achieves near-optimal privacy with EER approximately 0.49 across seven European languages

"our system achieves near-optimal privacy (EER $\approx$ 0.49), competitive intelligibility, and substantially better speech quality than dedicated anonymization baselines, while requiring no language-specific training."

Repurposing a multilingual voice cloning model for speaker anonymization produces substantially better speech quality than dedicated anonymization baselines without requiring language-specific training

"substantially better speech quality than dedicated anonymization baselines, while requiring no language-specific training"

Physics-integrated 3D Gaussian representations can now reconstruct deformable objects that can be simulated and rendered under explicit material models

"Physics-integrated 3D Gaussian representations now allow reconstructed deformable objects to be simulated and rendered under explicit material models."

KnockGS can estimate elasticity and density scales of 3D Gaussian objects from their dynamics under known applied forces

"We propose KnockGS, an interaction-response PhysicalGS framework that estimates the elasticity and density scales of a 3D Gaussian object from its dynamics under a known applied force."

KnockGS recovers material scales substantially more accurately than response retrieval, global regression, or fixed default materials across five held-out material targets

"Across five held-out material targets, our method recovers the scales substantially more accurately than response retrieval, global regression, or a fixed default material"