Introspective Cross-Attention Probing for Lightweight Transfer of Pre-trained Models

We propose InCA, a lightweight method for transfer learning that cross-attends to any activation layer of a pre-trained model. During training, InCA uses a single forward pass to extract multiple activations, which are passed to external cross-attention adapters, trained anew and combined or selected for downstream tasks. We show that, even when selecting a single top-scoring adapter, InCA achieves performance comparable to full fine-tuning, at a cost comparable to fine-tuning just the last layer. For example, with a cross-attention probe 1.3% the size of a pre-trained ViT-L/16 model, we achieve performance within 0.2% of the full fine-tuning paragon at 51% training cost of the baseline, on average across 11 downstream classification tasks. Unlike other forms of efficient adaptation, InCA does not require backpropagating through the pre-trained model, thus leaving its execution unaltered at both training and inference. The versatility of InCA is best illustrated in fine-grained tasks, which may require accessing information absent in the last layer but accessible in intermediate layer activations. Since the backbone is fixed, InCA allows parallel ensembling as well as parallel execution of multiple tasks. InCA achieves state-of-the-art performance in the ImageNet-to-Sketch multi-task benchmark.

The Best of Both Worlds: Combining Model-based and Nonparametric Approaches for 3D Human Body Estimation

Nonparametric based methods have recently shown promising results in reconstructing human bodies from monocular images while model-based methods can help correct these estimates and improve prediction. However, estimating model parameters from global image features may lead to noticeable misalignment between the estimated meshes and image evidence. To address this issue and leverage the best of both worlds, we propose a framework of three consecutive modules. A dense map prediction module explicitly establishes the dense UV correspondence between the image evidence and each part of the body model. The inverse kinematics module refines the key point prediction and generates a posed template mesh. Finally, a UV inpainting module relies on the corresponding feature, prediction and the posed template, and completes the predictions of occluded body shape. Our framework leverages the best of non-parametric and model-based methods and is also robust to partial occlusion. Experiments demonstrate that our framework outperforms existing 3D human estimation methods on multiple public benchmarks.

Task Adaptive Parameter Sharing for Multi-Task Learning

Adapting pre-trained models with broad capabilities has become standard practice for learning a wide range of downstream tasks. The typical approach of fine-tuning different models for each task is performant, but incurs a substantial memory cost. To efficiently learn multiple downstream tasks we introduce Task Adaptive Parameter Sharing (TAPS), a general method for tuning a base model to a new task by adaptively modifying a small, task-specific subset of layers. This enables multi-task learning while minimizing resources used and competition between tasks. TAPS solves a joint optimization problem which determines which layers to share with the base model and the value of the task-specific weights. Further, a sparsity penalty on the number of active layers encourages weight sharing with the base model. Compared to other methods, TAPS retains high accuracy on downstream tasks while introducing few task-specific parameters. Moreover, TAPS is agnostic to the model architecture and requires only minor changes to the training scheme. We evaluate our method on a suite of fine-tuning tasks and architectures (ResNet, DenseNet, ViT) and show that it achieves state-of-the-art performance while being simple to implement.

GeoFill: Reference-Based Image Inpainting of Scenes with Complex Geometry

Reference-guided image inpainting restores image pixels by leveraging the content from another reference image. The previous state-of-the-art, TransFill, warps the source image with multiple homographies, and fuses them together for hole filling. Inspired by structure from motion pipelines and recent progress in monocular depth estimation, we propose a more principled approach that does not require heuristic planar assumptions. We leverage a monocular depth estimate and predict relative pose between cameras, then align the reference image to the target by a differentiable 3D reprojection and a joint optimization of relative pose and depth map scale and offset. Our approach achieves state-of-the-art performance on both RealEstate10K and MannequinChallenge dataset with large baselines, complex geometry and extreme camera motions. We experimentally verify our approach is also better at handling large holes.

Representation Consolidation for Training Expert Students

Traditionally, distillation has been used to train a student model to emulate the input/output functionality of a teacher. A more useful goal than emulation, yet under-explored, is for the student to learn feature representations that transfer well to future tasks. However, we observe that standard distillation of task-specific teachers actually *reduces* the transferability of student representations to downstream tasks. We show that a multi-head, multi-task distillation method using an unlabeled proxy dataset and a generalist teacher is sufficient to consolidate representations from task-specific teacher(s) and improve downstream performance, outperforming the teacher(s) and the strong baseline of ImageNet pretrained features. Our method can also combine the representational knowledge of multiple teachers trained on one or multiple domains into a single model, whose representation is improved on all teachers' domain(s).

Unsupervised Action Segmentation with Self-supervised Feature Learning and Co-occurrence Parsing

Temporal action segmentation is a task to classify each frame in the video with an action label. However, it is quite expensive to annotate every frame in a large corpus of videos to construct a comprehensive supervised training dataset. Thus in this work we explore a self-supervised method that operates on a corpus of unlabeled videos and predicts a likely set of temporal segments across the videos. To do this we leverage self-supervised video classification approaches to perform unsupervised feature extraction. On top of these features we develop CAP, a novel co-occurrence action parsing algorithm that can not only capture the correlation among sub-actions underlying the structure of activities, but also estimate the temporal trajectory of the sub-actions in an accurate and general way. We evaluate on both classic datasets (Breakfast, 50Salads) and emerging fine-grained action datasets (FineGym) with more complex activity structures and similar sub-actions. Results show that our method achieves state-of-the-art performance on all three datasets with up to 22\% improvement, and can even outperform some weakly-supervised approaches, demonstrating its effectiveness and generalizability.

A linearized framework and a new benchmark for model selection for fine-tuning

Fine-tuning from a collection of models pre-trained on different domains (a "model zoo") is emerging as a technique to improve test accuracy in the low-data regime. However, model selection, i.e. how to pre-select the right model to fine-tune from a model zoo without performing any training, remains an open topic. We use a linearized framework to approximate fine-tuning, and introduce two new baselines for model selection -- Label-Gradient and Label-Feature Correlation. Since all model selection algorithms in the literature have been tested on different use-cases and never compared directly, we introduce a new comprehensive benchmark for model selection comprising of: i) A model zoo of single and multi-domain models, and ii) Many target tasks. Our benchmark highlights accuracy gain with model zoo compared to fine-tuning Imagenet models. We show our model selection baseline can select optimal models to fine-tune in few selections and has the highest ranking correlation to fine-tuning accuracy compared to existing algorithms.

Exponential Moving Average Normalization for Self-supervised and Semi-supervised Learning

We present a plug-in replacement for batch normalization (BN) called exponential moving average normalization (EMAN), which improves the performance of existing student-teacher based self- and semi-supervised learning techniques. Unlike the standard BN, where the statistics are computed within each batch, EMAN, used in the teacher, updates its statistics by exponential moving average from the BN statistics of the student. This design reduces the intrinsic cross-sample dependency of BN and enhance the generalization of the teacher. EMAN improves strong baselines for self-supervised learning by 4-6/1-2 points and semi-supervised learning by about 7/2 points, when 1%/10% supervised labels are available on ImageNet. These improvements are consistent across methods, network architectures, training duration, and datasets, demonstrating the general effectiveness of this technique.

When Perspective Comes for Free: Improving Depth Prediction with Camera Pose Encoding

Monocular depth prediction is a highly underdetermined problem and recent progress has relied on high-capacity CNNs to effectively learn scene statistics that disambiguate estimation. However, we observe that such models are strongly biased by the distribution of camera poses seen during training and fail to generalize to novel viewpoints, even when the scene geometry distribution remains fixed. To address this challenge, we propose a factored approach that estimates pose first, followed by a conditional depth estimation model that takes an encoding of the camera pose prior (CPP) as input. In many applications, a strong test-time pose prior comes for free, e.g., from inertial sensors or static camera deployment. A factored approach also allows for adapting pose prior estimation to new test domains using only pose supervision, without the need for collecting expensive ground-truth depth required for end-to-end training. We evaluate our pose-conditional depth predictor (trained on synthetic indoor scenes) on a real-world test set. Our factored approach, which only requires camera pose supervision for training, outperforms recent state-of-the-art methods trained with full scene depth supervision on 10x more data.