Alexandre Morgand, PhD
@alexmrgd.bsky.social
Computer Vision Research Scientist at @Simulon , music lover, fond of scientific/musical/geeky/useless stuff
"No Pose at All Self-Supervised Pose-Free 3DGS from Sparse Views"
TLDR: 3DGS + no poses during training/inference; shared feature extraction backbone; simultaneous prediction of 3D Gaussian primitives+camera poses in a canonical space from unposed (1 feed-forward step).
TLDR: 3DGS + no poses during training/inference; shared feature extraction backbone; simultaneous prediction of 3D Gaussian primitives+camera poses in a canonical space from unposed (1 feed-forward step).
August 7, 2025 at 3:49 PM
"No Pose at All Self-Supervised Pose-Free 3DGS from Sparse Views"
TLDR: 3DGS + no poses during training/inference; shared feature extraction backbone; simultaneous prediction of 3D Gaussian primitives+camera poses in a canonical space from unposed (1 feed-forward step).
TLDR: 3DGS + no poses during training/inference; shared feature extraction backbone; simultaneous prediction of 3D Gaussian primitives+camera poses in a canonical space from unposed (1 feed-forward step).
"Any-to-Bokeh: One-Step Video Bokeh via Multi-Plane Image Guided Diffusion"
📖TL;DR: Any-to-Bokeh is a novel one-step video bokeh framework that converts arbitrary input videos into temporally coherent, depth-aware bokeh effects.
📖TL;DR: Any-to-Bokeh is a novel one-step video bokeh framework that converts arbitrary input videos into temporally coherent, depth-aware bokeh effects.
June 13, 2025 at 1:43 PM
"Any-to-Bokeh: One-Step Video Bokeh via Multi-Plane Image Guided Diffusion"
📖TL;DR: Any-to-Bokeh is a novel one-step video bokeh framework that converts arbitrary input videos into temporally coherent, depth-aware bokeh effects.
📖TL;DR: Any-to-Bokeh is a novel one-step video bokeh framework that converts arbitrary input videos into temporally coherent, depth-aware bokeh effects.
"QUEEN: QUantized Efficient ENcoding of Dynamic Gaussians for Streaming Free-viewpoint Videos"
TL;DR: Streamable free-viewpoint videos efficient representations for with dynamic Gaussians. Reduce model size to just 0.7 MB per frame while training in < 5s and rendering at 350 FPS
TL;DR: Streamable free-viewpoint videos efficient representations for with dynamic Gaussians. Reduce model size to just 0.7 MB per frame while training in < 5s and rendering at 350 FPS
June 11, 2025 at 9:13 AM
"QUEEN: QUantized Efficient ENcoding of Dynamic Gaussians for Streaming Free-viewpoint Videos"
TL;DR: Streamable free-viewpoint videos efficient representations for with dynamic Gaussians. Reduce model size to just 0.7 MB per frame while training in < 5s and rendering at 350 FPS
TL;DR: Streamable free-viewpoint videos efficient representations for with dynamic Gaussians. Reduce model size to just 0.7 MB per frame while training in < 5s and rendering at 350 FPS
"STORM: Spatio-Temporal Reconstruction Model for Large-Scale Outdoor Scenes"
TL;DR: Data driven transformer in a feed forward manner; dense reconstruction in dynamic environment with 3D gaussians and velocities; self-supervised scene flows
TL;DR: Data driven transformer in a feed forward manner; dense reconstruction in dynamic environment with 3D gaussians and velocities; self-supervised scene flows
May 20, 2025 at 4:49 PM
"STORM: Spatio-Temporal Reconstruction Model for Large-Scale Outdoor Scenes"
TL;DR: Data driven transformer in a feed forward manner; dense reconstruction in dynamic environment with 3D gaussians and velocities; self-supervised scene flows
TL;DR: Data driven transformer in a feed forward manner; dense reconstruction in dynamic environment with 3D gaussians and velocities; self-supervised scene flows
St4RTrack: Simultaneous 4D Reconstruction and Tracking in the World
TL;DR: a feed-forward; (reconstructs+tracks dynamic video content); dust3r-like pointmaps for a pair of frames captured at different moments (1/2)
TL;DR: a feed-forward; (reconstructs+tracks dynamic video content); dust3r-like pointmaps for a pair of frames captured at different moments (1/2)
April 22, 2025 at 4:30 PM
St4RTrack: Simultaneous 4D Reconstruction and Tracking in the World
TL;DR: a feed-forward; (reconstructs+tracks dynamic video content); dust3r-like pointmaps for a pair of frames captured at different moments (1/2)
TL;DR: a feed-forward; (reconstructs+tracks dynamic video content); dust3r-like pointmaps for a pair of frames captured at different moments (1/2)
FLARE: Feed-forward Geometry, Appearance and Camera Estimation from Uncalibrated Sparse Views
TL;DR: feed-forward model; cascaded learning paradigm with camera pose serving as the critical bridge, recognizing its essential role in mapping 3D structures onto 2D image planes.
TL;DR: feed-forward model; cascaded learning paradigm with camera pose serving as the critical bridge, recognizing its essential role in mapping 3D structures onto 2D image planes.
March 14, 2025 at 10:21 AM
FLARE: Feed-forward Geometry, Appearance and Camera Estimation from Uncalibrated Sparse Views
TL;DR: feed-forward model; cascaded learning paradigm with camera pose serving as the critical bridge, recognizing its essential role in mapping 3D structures onto 2D image planes.
TL;DR: feed-forward model; cascaded learning paradigm with camera pose serving as the critical bridge, recognizing its essential role in mapping 3D structures onto 2D image planes.
⚡️Fast3R: Towards 3D Reconstruction of 1000+ Images in One Forward Pass
TL;DR: multi-view generalization to DUSt3R; processing many views in parallel: Transformer-based architecture forwards N images in a single forward pass, bypassing the need for iterative alignment.
TL;DR: multi-view generalization to DUSt3R; processing many views in parallel: Transformer-based architecture forwards N images in a single forward pass, bypassing the need for iterative alignment.
March 13, 2025 at 10:07 AM
⚡️Fast3R: Towards 3D Reconstruction of 1000+ Images in One Forward Pass
TL;DR: multi-view generalization to DUSt3R; processing many views in parallel: Transformer-based architecture forwards N images in a single forward pass, bypassing the need for iterative alignment.
TL;DR: multi-view generalization to DUSt3R; processing many views in parallel: Transformer-based architecture forwards N images in a single forward pass, bypassing the need for iterative alignment.
🪄 VACE: All-in-One Video Creation and Editing
from @alibabagroup.bsky.social's Tongyi Lab with:
Zeyinzi Jiang* Zhen Han* Chaojie Mao*† Jingfeng Zhang Yulin Pan Yu Liu
*Equal contribution, †Project lead
from @alibabagroup.bsky.social's Tongyi Lab with:
Zeyinzi Jiang* Zhen Han* Chaojie Mao*† Jingfeng Zhang Yulin Pan Yu Liu
*Equal contribution, †Project lead
March 12, 2025 at 8:39 AM
🪄 VACE: All-in-One Video Creation and Editing
from @alibabagroup.bsky.social's Tongyi Lab with:
Zeyinzi Jiang* Zhen Han* Chaojie Mao*† Jingfeng Zhang Yulin Pan Yu Liu
*Equal contribution, †Project lead
from @alibabagroup.bsky.social's Tongyi Lab with:
Zeyinzi Jiang* Zhen Han* Chaojie Mao*† Jingfeng Zhang Yulin Pan Yu Liu
*Equal contribution, †Project lead
Difix3D+: Improving 3D Reconstructions with Single-Step Diffusion Models
TL;DR: single-step diffusion models; a single-step image diffusion model trained to enhance and remove artifacts in rendered novel views caused by underconstrained regions of the 3D representation.
TL;DR: single-step diffusion models; a single-step image diffusion model trained to enhance and remove artifacts in rendered novel views caused by underconstrained regions of the 3D representation.
March 10, 2025 at 8:43 AM
Difix3D+: Improving 3D Reconstructions with Single-Step Diffusion Models
TL;DR: single-step diffusion models; a single-step image diffusion model trained to enhance and remove artifacts in rendered novel views caused by underconstrained regions of the 3D representation.
TL;DR: single-step diffusion models; a single-step image diffusion model trained to enhance and remove artifacts in rendered novel views caused by underconstrained regions of the 3D representation.
A Distractor-Aware Memory (DAM) for Visual Object Tracking with SAM2
TL;DR: SAM2.1 based; distractor-distilled (DiDi) dataset to better study the distractor problem
TL;DR: SAM2.1 based; distractor-distilled (DiDi) dataset to better study the distractor problem
March 5, 2025 at 8:48 AM
A Distractor-Aware Memory (DAM) for Visual Object Tracking with SAM2
TL;DR: SAM2.1 based; distractor-distilled (DiDi) dataset to better study the distractor problem
TL;DR: SAM2.1 based; distractor-distilled (DiDi) dataset to better study the distractor problem
CAST: Component-Aligned 3D Scene Reconstruction from an RGB Image
TL;DR: object-level 2D segmentation+relative depth; GPT-based model to analyze inter-object spatial relationships; occlusion-aware large-scale 3D generation model
TL;DR: object-level 2D segmentation+relative depth; GPT-based model to analyze inter-object spatial relationships; occlusion-aware large-scale 3D generation model
March 4, 2025 at 8:41 AM
CAST: Component-Aligned 3D Scene Reconstruction from an RGB Image
TL;DR: object-level 2D segmentation+relative depth; GPT-based model to analyze inter-object spatial relationships; occlusion-aware large-scale 3D generation model
TL;DR: object-level 2D segmentation+relative depth; GPT-based model to analyze inter-object spatial relationships; occlusion-aware large-scale 3D generation model
Are diffusion models falling for optical illusion?
"The Art of Deception: Color Visual Illusions and Diffusion Models"
TL;DR: Diffusion models exhibit human-like perceptual shifts in brightness and color within their latent space.
"The Art of Deception: Color Visual Illusions and Diffusion Models"
TL;DR: Diffusion models exhibit human-like perceptual shifts in brightness and color within their latent space.
February 28, 2025 at 3:15 PM
Are diffusion models falling for optical illusion?
"The Art of Deception: Color Visual Illusions and Diffusion Models"
TL;DR: Diffusion models exhibit human-like perceptual shifts in brightness and color within their latent space.
"The Art of Deception: Color Visual Illusions and Diffusion Models"
TL;DR: Diffusion models exhibit human-like perceptual shifts in brightness and color within their latent space.
Does 3D Gaussian Splatting Need Accurate Volumetric Rendering?
TL;DR: While more accurate volumetric rendering can help for low numbers of primitives, efficient optimization + large number of Gaussians allows 3DGS to outperform volumetric rendering despite its approximations
TL;DR: While more accurate volumetric rendering can help for low numbers of primitives, efficient optimization + large number of Gaussians allows 3DGS to outperform volumetric rendering despite its approximations
February 27, 2025 at 9:58 AM
Does 3D Gaussian Splatting Need Accurate Volumetric Rendering?
TL;DR: While more accurate volumetric rendering can help for low numbers of primitives, efficient optimization + large number of Gaussians allows 3DGS to outperform volumetric rendering despite its approximations
TL;DR: While more accurate volumetric rendering can help for low numbers of primitives, efficient optimization + large number of Gaussians allows 3DGS to outperform volumetric rendering despite its approximations
The NeRF-life vengeance?
"Sparse Voxels Rasterization: Real-time High-fidelity Radiance Field Rendering"
"Sparse Voxels Rasterization: Real-time High-fidelity Radiance Field Rendering"
February 24, 2025 at 10:09 AM
The NeRF-life vengeance?
"Sparse Voxels Rasterization: Real-time High-fidelity Radiance Field Rendering"
"Sparse Voxels Rasterization: Real-time High-fidelity Radiance Field Rendering"
Self-Calibrating Gaussian Splatting for Large Field of View Reconstruction
TL;DR: Self calibration + cubemap-based resampling strategy to support large FOV images
TL;DR: Self calibration + cubemap-based resampling strategy to support large FOV images
February 20, 2025 at 8:50 AM
Self-Calibrating Gaussian Splatting for Large Field of View Reconstruction
TL;DR: Self calibration + cubemap-based resampling strategy to support large FOV images
TL;DR: Self calibration + cubemap-based resampling strategy to support large FOV images
Animate Anyone 2: High-Fidelity Character Image Animation with Environment Affordance
TL;DR: motion from source video + capture environmental representations as conditional inputs. Shape-agnostic mask strategy for character/environment relationship .
TL;DR: motion from source video + capture environmental representations as conditional inputs. Shape-agnostic mask strategy for character/environment relationship .
February 18, 2025 at 4:26 PM
Animate Anyone 2: High-Fidelity Character Image Animation with Environment Affordance
TL;DR: motion from source video + capture environmental representations as conditional inputs. Shape-agnostic mask strategy for character/environment relationship .
TL;DR: motion from source video + capture environmental representations as conditional inputs. Shape-agnostic mask strategy for character/environment relationship .
Pippo : High-Resolution Multi-View Humans from a Single Image
TL;DR: 1K Multiview Diffusion Transformer pre-trained on 3B Human images without captions; post-trained on 2.5K studio captures with pixel-aligned control via ControlMLP; generates > 5x views at inference
TL;DR: 1K Multiview Diffusion Transformer pre-trained on 3B Human images without captions; post-trained on 2.5K studio captures with pixel-aligned control via ControlMLP; generates > 5x views at inference
February 18, 2025 at 10:16 AM
Pippo : High-Resolution Multi-View Humans from a Single Image
TL;DR: 1K Multiview Diffusion Transformer pre-trained on 3B Human images without captions; post-trained on 2.5K studio captures with pixel-aligned control via ControlMLP; generates > 5x views at inference
TL;DR: 1K Multiview Diffusion Transformer pre-trained on 3B Human images without captions; post-trained on 2.5K studio captures with pixel-aligned control via ControlMLP; generates > 5x views at inference
Since 2024, it's crazy how competitive the field of generative video is. Here is another player but open source this time!
Hong Kong University and ByteDance present "Goku: Flow Based Video Generative Foundation Models"
Hong Kong University and ByteDance present "Goku: Flow Based Video Generative Foundation Models"
February 14, 2025 at 9:21 AM
Since 2024, it's crazy how competitive the field of generative video is. Here is another player but open source this time!
Hong Kong University and ByteDance present "Goku: Flow Based Video Generative Foundation Models"
Hong Kong University and ByteDance present "Goku: Flow Based Video Generative Foundation Models"
📜 Fillerbuster: Multi-View Scene Completion for Casual Captures
TL;DR: Unified framework for scene completion; joint models images and camera poses estimation to reconstruct missing parts of casually captured scenes. 1B-parameter diffusion model from scratch.
TL;DR: Unified framework for scene completion; joint models images and camera poses estimation to reconstruct missing parts of casually captured scenes. 1B-parameter diffusion model from scratch.
February 12, 2025 at 9:02 AM
📜 Fillerbuster: Multi-View Scene Completion for Casual Captures
TL;DR: Unified framework for scene completion; joint models images and camera poses estimation to reconstruct missing parts of casually captured scenes. 1B-parameter diffusion model from scratch.
TL;DR: Unified framework for scene completion; joint models images and camera poses estimation to reconstruct missing parts of casually captured scenes. 1B-parameter diffusion model from scratch.
Diffusion as Shader: 3D-aware Video Diffusion for Versatile Video Generation Control
TL;DR: Manipulating 3D tracking videos; link frames, significantly enhancing for temporal consistency of the generated videos; 3 days oftraining on 8 H800 GPUs using less than 10k videos
TL;DR: Manipulating 3D tracking videos; link frames, significantly enhancing for temporal consistency of the generated videos; 3 days oftraining on 8 H800 GPUs using less than 10k videos
February 11, 2025 at 8:42 AM
Diffusion as Shader: 3D-aware Video Diffusion for Versatile Video Generation Control
TL;DR: Manipulating 3D tracking videos; link frames, significantly enhancing for temporal consistency of the generated videos; 3 days oftraining on 8 H800 GPUs using less than 10k videos
TL;DR: Manipulating 3D tracking videos; link frames, significantly enhancing for temporal consistency of the generated videos; 3 days oftraining on 8 H800 GPUs using less than 10k videos
Zero-Shot Novel View and Depth Synthesis with Multi-View Geometric Diffusion
TL;DR: diffusion-based; raymap conditioning to both augment visual features with spatial information from different viewpoints; multi-task generation of images and depth maps
TL;DR: diffusion-based; raymap conditioning to both augment visual features with spatial information from different viewpoints; multi-task generation of images and depth maps
February 4, 2025 at 9:47 AM
Zero-Shot Novel View and Depth Synthesis with Multi-View Geometric Diffusion
TL;DR: diffusion-based; raymap conditioning to both augment visual features with spatial information from different viewpoints; multi-task generation of images and depth maps
TL;DR: diffusion-based; raymap conditioning to both augment visual features with spatial information from different viewpoints; multi-task generation of images and depth maps
DiffVSR Enhancing Real-World Video Super-Resolution with Diffusion Models for Advanced Visual Quality and Temporal Consistency
TL;DR: multi-scale temporal attention module for spatial accuracy. Noise rescheduling mechanism & latent transition approach for temporal consistency
TL;DR: multi-scale temporal attention module for spatial accuracy. Noise rescheduling mechanism & latent transition approach for temporal consistency
February 3, 2025 at 11:10 AM
DiffVSR Enhancing Real-World Video Super-Resolution with Diffusion Models for Advanced Visual Quality and Temporal Consistency
TL;DR: multi-scale temporal attention module for spatial accuracy. Noise rescheduling mechanism & latent transition approach for temporal consistency
TL;DR: multi-scale temporal attention module for spatial accuracy. Noise rescheduling mechanism & latent transition approach for temporal consistency
CubeDiff: Repurposing Diffusion-Based Image Models for Panorama Generation
TL;DR: 360° panoramas using diffusion-based image models. cubemap representations + fine-tuning pretrained txt2img models, CubeDiff simplifies the panorama generation process, delivering high-quality, consistent panoramas.
TL;DR: 360° panoramas using diffusion-based image models. cubemap representations + fine-tuning pretrained txt2img models, CubeDiff simplifies the panorama generation process, delivering high-quality, consistent panoramas.
January 31, 2025 at 10:03 AM
CubeDiff: Repurposing Diffusion-Based Image Models for Panorama Generation
TL;DR: 360° panoramas using diffusion-based image models. cubemap representations + fine-tuning pretrained txt2img models, CubeDiff simplifies the panorama generation process, delivering high-quality, consistent panoramas.
TL;DR: 360° panoramas using diffusion-based image models. cubemap representations + fine-tuning pretrained txt2img models, CubeDiff simplifies the panorama generation process, delivering high-quality, consistent panoramas.
BLADE: Single-view Body Mesh Learning through Accurate Depth Estimation
TL;DR: fully perspective projection model without applying heuristics; depth, focal parameters, 3D pose, and 2D alignment estimation
TL;DR: fully perspective projection model without applying heuristics; depth, focal parameters, 3D pose, and 2D alignment estimation
January 29, 2025 at 8:25 AM
BLADE: Single-view Body Mesh Learning through Accurate Depth Estimation
TL;DR: fully perspective projection model without applying heuristics; depth, focal parameters, 3D pose, and 2D alignment estimation
TL;DR: fully perspective projection model without applying heuristics; depth, focal parameters, 3D pose, and 2D alignment estimation
Continuous 3D Perception Model with Persistent State
TL;DR: An online 3D reasoning framework for various 3D tasks from only RGB inputs
TL;DR: An online 3D reasoning framework for various 3D tasks from only RGB inputs
January 27, 2025 at 9:37 AM
Continuous 3D Perception Model with Persistent State
TL;DR: An online 3D reasoning framework for various 3D tasks from only RGB inputs
TL;DR: An online 3D reasoning framework for various 3D tasks from only RGB inputs