Teaser

The Visual Computing Laboratory (VCL) paper reading group is a forum for reading, presenting and discussing papers related to computer graphics, image processing, machine learning, computer vision and visualization. We meet on a biweekly basis and discuss a selected paper related to these areas.

Participants are welcome to join based on time and interest. However, there is also the option to participate in the reading group as part of a PhD course and collect course credits. Information on the examination is detailed below.

General information


Aim

The aim of the reading group is to both learn about published research and to practice reading and discussing research. More specifically, the aim is to:

  • keep up-to-date with recent research. This goes both for topics related to your research, but also those not directly related. There are many intersections between computer graphics, image processing, machine learning, computer vision and visualization, and a broadened perspective will both increase knowledge and open up for future ideas.
  • practice reading papers.
  • improve paper writing and reviewing skills by identifying good and bad ways of presenting research in paper form, i.e. learn-by-example.
  • discuss relevant research.
  • inspire and communicate ideas. Many ideas are born from reading and discussing research, not least in cross-disciplinary applications, e.g. "how can the method in area A be used in area B?".

Structure

For each paper reading seminar, one person, the presenter, selects a paper in advance which all of the participants reads before the seminar. The presenter for an upcoming seminar is either selected when we meet, or by notifying me through email. During a reading seminar, the presenter gives a short (10-15 min) conference-style presentation of the paper. The presentation is followed by a short (5-10 min) review, which is supposed to comment on:

  • the impact of the work, i.e. the novelty and usefulness of the work.
  • paper presentation, both positive and negative sides (e.g. structure, writing, figures, etc.).
  • possible extensions or applications related to own research.

Finally, the presenter opens up a discussion around the paper by providing a few example discussion points. The discussion (20-30 min) expects participants to actively take part in the discussions. This relies on that the participants have read the paper before the meeting.

Examination

Participation in the reading group can be based on time and interest, e.g. only taking part in seminars which are deemed most relevant to your research. There is no need to register for this, but in order to be included on the email list, please let me know by email. However, in order to collect course credits you are required to present and actively participate in discussions. For each course credit (1hp), you are required to

  • be the presenter of one reading seminar.
  • actively participate in a total of 3 seminars.

For example, by giving 2 presentations and participating in 6 seminars (including the ones as presenter), you will be awarded 2hp.

In order to sign up for the paper reading PhD course, let me know in advance through email: gabriel.eilertsen@liu.se.

Schedule

Dates are tentative and subject to change, but the intention is to aim for Tuesdays at 10.15 every other week. Depending on the number of participants, there is also the option to have two papers in a longer seminar session.


DatePresenterPaperKeywords
Aug. 29, 10.15-11.00BehnazZhou et al. 2023 PhotoMat: A Material Generator Learned from Single Flash PhotosBRDF, material measurement, machine learning, deep learning, generative modelling
Oct. 10, 10.15-11.00Gabriel B.Oord et al. 2017, Neural Discrete Representation Learning
Esser et al. 2021, Taming Transformers for High-Resolution Image Synthesis 		vector quantization, machine learning, deep learning, generative modelling, VAEs, GANs
Oct. 24, 10.15-11.00Saghi
Nov. 07, 10.15-11.00TBA
Nov. 21, 10.15-11.00TBA
Dec. 05, 10.15-11.00TBA
Dec. 19, 10.15-11.00TBA


Paper log:

DatePresenterPaperKeywords
Jan. 31, 10.15-11.00Watching and discussing NeurIPS 2022 pre-recorded presentations.
Feb. 15, 10.15-11.00ArtyRombach et al. 2022, High-Resolution Image Synthesis With Latent Diffusion Modelsmachine learning, deep learning, generative modelling, diffusion models
Apr. 11, 10.15-11.00WenGuo et al. 2020, MaterialGAN: reflectance capture using a generative SVBRDF modelBRDF, material measurement, machine learning, deep learning, generative modelling
May 24, 13.15-14.00SaghiPresentation and discussion of interesting papers from EG2023

DatePresenterPaperKeywords
Sep. 06, 10.15-11.00SaghiMüller et al. 2022, Instant Neural Graphics Primitives with a Multiresolution Hash Encodingmachine learning, deep learning, scene representation, view synthesis, neural representation
Sep. 19, 10.00-11.00Watching and discussing SIGGRAPH 2022 pre-recorded presentations.
Oct. 04, 10.15-11.00Gabriel B.Kellnhofer et al. 2021, Neural Lumigraph Renderingmachine learning, deep learning, scene representation, view synthesis, neural representation
Oct. 18, 10.15-11.00BehnazLagunas et al. 2019, A similarity measure for material appearancematerial appearance, machine learning, deep learning, perception
Nov. 15, 10.15-11.00WenZhang et al. 2020, Optimization-Inspired Compact Deep Compressive Sensingmachine learning, deep learning, compressed sensing, image reconstruction
Nov. 29, 10.15-11.00YifanPeebles et al. 2022, GAN-Supervised Dense Visual Alignmentmachine learning, deep learning, generative modelling, visual alignment

DatePresenterPaperKeywords
Jan. 25, 10.15-11.00BehnazRainer et al. 2019, Neural BTF Compression and Interpolationimage-based rendering, machine learning, deep learning, image compression
Feb. 8, 10.15-11.00Gabriel B.Zhang et al. 2018, The Unreasonable Effectiveness of Deep Features as a Perceptual Metricimage quality assessment, IQA, image comparison, image metric, machine learning, deep learning
Mar. 8, 10.15-11.00EhsanWu et al. 2019, Learning a Compressed Sensing Measurement Matrix via Gradient Unrollingmachine learning, deep learning, compressed sensing
Mar. 22, 10.15-11.00WenMatusik et al. 2003, A data-driven reflectance modellight reflection models, photometric measurements, reflectance, BRDF, image-based modeling
Apr. 5, 10.15-11.00TanaboonSztrajman et al. 2021, Neural BRDF Representation and Importance Samplinglight reflection models, photometric measurements, BRDF, machine learning, deep learning
May. 3, 10.15-11.00BehnazWang et al. 2019, HyperReconNet: Joint Coded Aperture Optimization and Image Reconstruction for Compressive Hyperspectral Imaginghyperspectral imaging, image reconstruction, image coding, compressive imaging, deep learning
May. 17, 10.15-11.00MildaWang et al. 2019, Aleatoric uncertainty estimation with test-time augmentation for medical image segmentation with convolutional neural networksuncertainty estimation, deep learning, medical imaging, test-time augmentations, image segmentation
May. 31, 10.15-11.00SaghiMizuno et al. 2022, Acquiring a Dynamic Light Field through a Single-Shot Coded Imagelight fields, computational photography, machine learning, deep learning, image coding

DatePresenterPaperKeywords
Aug. 31, 10.15-11.00SaghiScetbon et al. 2021, Deep K-SVD Denoisingnoise reduction, machine learning, image processing
Sep. 15, 10.15-11.00MildaKolesnikov et al. 2020, Big Transfer (BiT): General Visual Representation Learningdeep learning, computer vision, image recognition
Oct. 12, 10.15-11.00RymYan et al. 2020, On Robustness of Neural Ordinary Differential Equationsneural ODE, deep learning, robustness
Oct. 26, 10.15-11.00KarinBirhane and Prabhu 2020, Large image datasets: A pyrrhic win for computer vision?deep learning, fairness, bias
Nov. 09, 10.15-11.00WenNielsen et al. 2015, On Optimal, Minimal BRDF Sampling for Reflectance Acquisitionreflectance, BRDF, MERL, reconstruction
Dec. 7, 10.15-11.00RymNovak et al. 2018, Sensitivity and Generalization in Neural Networks: an Empirical Studydeep learning, sensitivity analysis, generalization

DatePresenterPaperKeywords
Jan. 26, 10.15-11.00Gabriel B.Sun et al. 2019, Single Image Portrait Relightingcomputer graphics, deep learning, image-based rendering, computational photography
Feb. 9, 10.15-11.00ApostoliaHou et al. 2019, Robust Histopathology Image Analysis: to Label or to Synthesize?
(builds on the earlier arXiv version: Unsupervised Histopathology Image Synthesis)		deep learning, digital pathology, image synthesis, generative learning, GANs
Feb. 23, 10.15-11.00SaghiSantos et al. 2020, Single Image HDR Reconstruction Using a CNN with Masked Features and Perceptual Lossdeep learning, image processing, high dynamic range imaging
Mar. 9, 10.15-11.00TanaboonTongbuasirilai et al. 2021, A Non-parametric Sparse BRDF Modelcomputer graphics, reflectance modeling, BRDF modeling, sparse representation, dictionary learning
Mar. 23, 10.15-11.00KarinBender et al. 2021, On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?deep learning, natural language processing, ethical AI
Apr. 6, 10.15-11.00Gabriel B.Nabati et al. 2018, Fast and Accurate Reconstruction of Compressed Color Light Fieldcomputational photography, light fields, machine learning, deep learning
Apr. 20, 10.15-11.00MildaDosovitskiy et al. 2021, An Image is Worth 16x16 Words: Transformers for Image Recognition at Scaledeep learning, computer vision, image recognition, self-attention, transformer
May 4, 10.15-11.00WenIordache et al. 2011, Sparse Unmixing of Hyperspectral Dataabundance estimation, convex optimization, hyperspectral imaging, sparse regression, spectral unmixing
May 18, 10.15-11.00Gabriel E.Tolstikhin et al. 2021, MLP-Mixer: An all-MLP Architecture for Visiondeep learning, computer vision, image recognition, MLP
Jun. 1, 10.15-11.00BehnazShi et al. 2019, Image Compressed Sensing Using Convolutional Neural Networkcompressed sensing, deep learning, convolutional neural network, sampling matrix, image reconstruction

DatePresenterPaperKeywords
Sep. 8, 10.15-11.00MildaKarras et al. 2020, Analyzing and Improving the Image Quality of StyleGAN
(first StyleGAN can be found in: A Style-Based Generator Architecture for Generative Adversarial Networks)		deep learning, unsupervised learning, generative learning, GANs
Sep. 23, 10.15-11.00KristoferMartin-Brualla et al. 2020, NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collections
(builds on: Representing Scenes as Neural Radiance Fields for View Synthesis)		scene representation, view synthesis, image-based rendering, volume rendering, 3D deep learning
Oct. 6, 10.15-11.00FereshtehMehta and Egiazarian 2016, Texture Classification Using Dense Micro-Block Differencetexture classification, descriptors, compressive sensing, LBP, SVM, Scale Invariant Feature Transform
Oct. 21, 10.15-11.00ApostoliaKarras et al. 2020, Training Generative Adversarial Networks with Limited Datadeep learning, unsupervised learning, generative learning, GANs
Nov. 3, 10.15-11.00MildaSchirrmeister et al. 2020, Understanding anomaly detection with deep invertible networks through hierarchies of distributions and features
(uses flow based generative models: Glow: Generative flow with invertible 1x1 convolutions)		deep learning, anomaly detection, unsupervised learning, generative learning, GLOW
Nov. 17, 10.15-11.00Gabriel E.Liu et al. 2020, Diverse Image Generation via Self-Conditioned GANsdeep learning, unsupervised learning, generative learning, GANs
Dec. 1, 10.15-11.00KristoferAnonymous, LambdaNetworks: Modeling long-range Interactions without Attention (ICLR 2021 submission)deep learning, neural networks, attention, transformer, vision, image classification
Dec. 15, 10.15-11.00KarinTaori et al. 2020, Measuring Robustness to Natural Distribution Shifts in Image Classificationdeep learning, machine learning, domain shift, generalization

DatePresenterPaperKeywords
Feb. 4, 10.15-11.00Tanaboon Deschaintre et al. 2019, Flexible SVBRDF Capture with a Multi-Image Deep Networkcomputer graphics, deep learning, material capturing, rendering
Feb. 18, 10.15-11.00KristoferUlyanov et al. 2018, Deep Image Prior (extended in "Double-DIP": Unsupervised Image Decomposition via Coupled Deep-Image-Priors)deep learning, computer vision, deep convolutional networks
Mar. 3, 10.15-11.00KarinOord et al. 2018, Representation Learning with Contrastive Predictive Codingdeep learning, representation learning, unsupervised learning
Mar. 17, 10.15-11.00WitoGünther et al. 2014, Opacity Optimization for Surfacesvisualization, integral surfaces, flow visualization, computer graphics
Mar. 31, 10.15-11.00ApostoliaXiangli et al. 2020, Real or Not Real, that is the Questiondeep learning, image synthesis, generative learning, GANs
Apr. 14, 10.15-12.00MildaSchlegl et al. 2017, Unsupervised Anomaly Detection with Generative Adversarial Networks to Guide Marker Discoverydeep learning, unsupervised learning, generative learning, GANs, medical imaging
ElmiraAlbo et al. 2016, Off the Radar: Comparative Evaluation of Radial Visualization Solutions for Composite Indicatorsvisualization, multi-dimensional visualization, visualization evaluation, radial layout design
Apr. 28, 10.15-11.00JensWilson et al. 2018, Evolving simple programs for playing Atari gamescomputer vision, genetic programming, artificial intelligence, reinforcement learning
May 12, 10.15-11.00Gabriel E.Frankle & Carbin 2019, The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks
[related papers]		deep learning, neural networks, sparsity, pruning, compression
May 26, 10.15-11.00Gabriel B.Inagaki et al. 2018, Learning to Capture Light Fields through a Coded Aperture Cameracomputational photography, light fields, machine learning, deep learning
Jun. 9, 10.15-12.00FereshtehGao et al. 2019, Deep Restoration of Vintage Photographs From Scanned Halftone Printsmachine learning, deep learning, image processing, deep convolutional networks
KristoferSchmidhuber 2019, Reinforcement Learning Upside Down: Don't Predict Rewards – Just Map Them to Actionsmachine learning, deep learning, reinforcement learning

Paper suggestions

Papers can be chosen from being closely related to your own research, i.e. papers you would read anyway. Relevant papers can also be found by browsing papers from recent conferences such as SIGGRAPH, SIGGRAPH Asia, Eurographics, Pacific graphics, CVPR, ICCV, ECCV, NeurIPS, ICML, ICLR, etc., or journals such as IEEE TIP, IEEE TPAMI, IEEE TVCG, ACM ToG, CGF, etc. Recent papers can also be found on open archives such as arXiv.

Papers should be related to computer graphics, image processing, machine learning, computer vision and/or visualization. Papers that present work in the intersection of some of these areas are perfect for the reading group. Papers can also present more fundamental techniques and ideas, which are applicable in the areas.

The following list will be continuously updated with paper suggestions. It will both contain papers that I would find interesting to read, as well as suggestions from other reading group participants. Please feel free to provide me with suggestions you find interesting.


AuthorsTitleResource
Richardson et al.Encoding in style: a stylegan encoder for image-to-image translation[CVPR 2021]
Park et al.Semantic image synthesis with spatially-adaptive normalization[CVPR 2019]
Dai et al.CoAtNet: Marrying Convolution and Attention for All Data Sizes[arXiv 2021]
Dhariwal & NicholDiffusion models beat GANs on image synthesis[arXiv 2021]
Queiruga et al.Stateful ODE-Nets using Basis Function Expansions[NeurIPS 2021]
Tian et al.Understanding Self-Supervised Learning Dynamics without Contrastive Pairs[arXiv 2021]
Tov et al.Designing an encoder for stylegan image manipulation[Siggraph 2021]
Hudson & ZitnickGenerative adversarial transformers[arXiv 2021]
Richter et al.Enhancing Photorealism Enhancement[arXiv 2021]
Preechakul et al.Diffusion Autoencoders: Toward a Meaningful and Decodable Representation[CVPR 2022]



Last updated:  2023-10-03, Gabriel Eilertsen