You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: _site/js/publications_data.js
+21-20Lines changed: 21 additions & 20 deletions
Original file line number
Diff line number
Diff line change
@@ -1,24 +1,5 @@
1
1
constPUBLICATIONS_DATA_LOCAL=[
2
-
{
3
-
"title": "Fitting Image Diffusion Models on Video Datasets",
4
-
"authors": [
5
-
"Juhun Lee",
6
-
"Simon S. Woo"
7
-
],
8
-
"venue_full": "Workshop on International Conference on Computer Vision",
9
-
"venue": "ICCV",
10
-
"track": "Workshop Paper",
11
-
"Factor": [
12
-
"",
13
-
0
14
-
],
15
-
"year": 2026,
16
-
"links": {
17
-
"conf": "https://iccv.thecvf.com/"
18
-
},
19
-
"img": "/img/Publications/2026_ICCVW_Juhun.png",
20
-
"abstract": "Image diffusion models are trained on independently sampled static images. While this is the bedrock task protocol in generative modeling, capturing the temporal world through the lens of static snapshots is information-deficient by design. This limitation leads to slower convergence, limited distributional coverage, and reduced generalization. In this work, we propose a simple and effective training strategy that leverages the temporal inductive bias present in continuous video frames to improve diffusion training. Notably, the proposed method requires no architectural modification and can be seamlessly integrated into standard diffusion training pipelines. We evaluate our method on the HandCo dataset, where hand-object interactions exhibit dense temporal coherence andsubtle variations in finger articulation often result in semantically distinct motions. Empirically, our method accelerates convergence by over 2x faster and achieves lower FID on both training and validation distributions. It also improves generative diversity by encouraging the model to capture meaningful temporal variations. We further provide an optimization analysis showing that our regularization reduces the gradient variance, which contributes to faster convergence."
21
-
},
2
+
22
3
{
23
4
"title": "ICR-NET: Robust Deepfake Detection under Temporal Corruption",
"abstract": "As pretrained models are increasingly shared on the web, ensuring that models can forget or delete sensitive, copyrighted, or private information upon request has become crucial. Machine unlearning has been proposed to address this issue. However, current evaluations for unlearning methods rely on output-based metrics, which cannot verify whether information is completely deleted or merely suppressed at the representation level, where suppression is insufficient for true unlearning. To address this gap, we propose a novel restoration-based analysis framework that uses Sparse Autoencoders to identify class-specific expert features in intermediate layers and applies inference-time steering to quantitatively distinguish between suppression and deletion. Applying our framework to 12 major unlearning methods in image classification tasks, we find that most methods achieve high restoration rates of unlearned information, indicating that they only suppress information at the decision-boundary level, while preserving semantic features in intermediate representations. Notably, even retraining from pretrained checkpoints shows high restoration, revealing that pretrained feature hierarchies persist. These results demonstrate that representation-level retention poses significant risks overlooked by output-based metrics, highlighting the need for new unlearning evaluation criteria. We propose new evaluation guidelines that prioritize representation-level verification, especially for privacy-critical applications in the pretrained model era."
113
94
},
95
+
{
96
+
"title": "Fitting Image Diffusion Models on Video Datasets",
97
+
"authors": [
98
+
"Juhun Lee",
99
+
"Simon S. Woo"
100
+
],
101
+
"venue_full": "Workshop on International Conference on Computer Vision",
102
+
"venue": "ICCV",
103
+
"track": "Workshop Paper",
104
+
"Factor": [
105
+
"",
106
+
0
107
+
],
108
+
"year": 2025,
109
+
"links": {
110
+
"conf": "https://iccv.thecvf.com/"
111
+
},
112
+
"img": "/img/Publications/2025_ICCVW_Juhun.png",
113
+
"abstract": "Image diffusion models are trained on independently sampled static images. While this is the bedrock task protocol in generative modeling, capturing the temporal world through the lens of static snapshots is information-deficient by design. This limitation leads to slower convergence, limited distributional coverage, and reduced generalization. In this work, we propose a simple and effective training strategy that leverages the temporal inductive bias present in continuous video frames to improve diffusion training. Notably, the proposed method requires no architectural modification and can be seamlessly integrated into standard diffusion training pipelines. We evaluate our method on the HandCo dataset, where hand-object interactions exhibit dense temporal coherence andsubtle variations in finger articulation often result in semantically distinct motions. Empirically, our method accelerates convergence by over 2x faster and achieves lower FID on both training and validation distributions. It also improves generative diversity by encouraging the model to capture meaningful temporal variations. We further provide an optimization analysis showing that our regularization reduces the gradient variance, which contributes to faster convergence."
114
+
},
114
115
{
115
116
"title": "Self-Disclosure of Mental Health via Deepfakes: Testing the Effects of Self-Deepfakes on Affective Resistance and Intention to Seek Mental Health Support",
0 commit comments