Canada

Reinforcement Learning from Demonstrations (RLfD) has emerged as an effective method by fusing expert demonstrations into Reinforcement Learning (RL) training, harnessing the strengths of both Imitation Learning (IL) and RL. However, existing algorithms rely on offline demonstrations, which can introduce a distribution gap between the demonstrations and the actual training environment, limiting their performance. In this paper, we propose a novel approach, Reinforcement Learning from Online Demonstrations (RLfOLD), that leverages online demonstrations to address this limitation, ensuring the agent learns from relevant and up-to-date scenarios, thus effectively bridging the distribution gap. Unlike conventional policy networks used in typical actor-critic algorithms, RLfOLD introduces a policy network that outputs two standard deviations: one for exploration and the other for IL training. This novel design allows the agent to adapt to varying levels of uncertainty inherent in both RL and IL. Furthermore, we introduce an exploration process guided by an online expert, incorporating an uncertainty-based technique. Our experiments on the CARLA NoCrash benchmark demonstrate the effectiveness and efficiency of RLfOLD. Notably, even with a significantly smaller encoder and a single camera setup, RLfOLD surpasses state-of-the-art methods in this evaluation. These results, achieved with limited resources, highlight RLfOLD as a highly promising solution for real-world applications.

AAAI 2024

RLfOLD: Reinforcement Learning from Online Demonstrations in Urban Autonomous Driving

autonomous driving

imitation learning

reinforcement learning

technical paper

We are pleased to announce the Thirty-Eighth AAAI Conference on Artificial Intelligence (AAAI-24), which will be held in Vancouver, British Columbia at the Vancouver Convention Centre – West Building from 20-27 February, 2024.

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. AAAI-24 will feature technical paper presentations, special tracks, invited speakers, workshops, tutorials, poster sessions, senior member presentations, competitions, and exhibit programs, and a range of other activities to be announced.

We expect for AAAI-24 to be an in-person conference – one author of all accepted papers will be expected to present work in person unless there are exceptional circumstances that prevent this.<br><br><br><br>

In order to access the AAAI-24 event page you need to register [here](https://aaai.org/aaai-conference/registration/)

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. 

Constrained Reinforcement Learning has been employed to compute safe policies through the use of trajectory based cost constraints (on expected cost or Value at Risk, VaR cost or Conditional VaR cost). The key challenge is in handling constraints on cost accumulated across time steps. Existing methods have developed innovative ways of converting such trajectory based cost constraint to surrogate RL problems that can be solved easily. While such approaches have provided good solutions with regards to objective, they rely on cost estimates and can therefore be either aggressive or conservative with respect to costs. This is owing to use of estimates for "future" or "backward" costs in local cost constraints.

To that end, we provide an equivalent unconstrained formulation to constrained RL that employs reward penalties over augmented states to compute safe policies. Due to being able to identify specific trajectories that violate cost constraints accurately (and not just in expectation), we are able to provide approaches which have strong empirical performance and theoretical properties. More importantly, this provides a new paradigm for solving richly constrained (e.g., constraints on expected cost, Value at Risk, Conditional Value at Risk) Reinforcement Learning problems effectively. As we show in our experimental results, we are able to outperform leading approaches for different constraint types on multiple benchmark problems.

Reward Penalties on Augmented States for Solving Richly Constrained RL Effectively

Diffusion-based generative models represent the current state-of-the-art for image generation. However, standard diffusion models are based on Euclidean geometry and do not translate directly to manifold-valued data. In this work, we develop extensions of both score-based generative models (SGMs) and Denoising Diffusion Probabilistic Models (DDPMs) to the Lie group of 3D rotations, SO(3). SO(3) is of particular interest in many disciplines such as robotics, biochemistry and astronomy/cosmology science. Contrary to more general Riemannian manifolds, SO(3) admits a tractable solution to heat diffusion, and allows us to implement efficient training of diffusion models. We apply both SO(3) DDPMs and SGMs to synthetic densities on SO(3) and demonstrate state-of-the-art results. Additionally, we demonstrate the practicality of our model on pose estimation tasks and in predicting correlated galaxy orientations for astrophysics/cosmology.

Unified Framework for Diffusion Generative Models in SO(3): Applications in Computer Vision and Astrophysics

Parity reasoning is challenging for Conflict-Driven Clause Learning (CDCL) SAT solvers. This has been observed even for simple formulas encoding two contradictory parity constraints with different variable orders (Chew and Heule 2020). We provide an analytical explanation for their hardness by showing that they require exponential resolution refutations with high probability when the variable order is chosen at random. We obtain this result by proving that these formulas, which are known to be Tseitin formulas, have Tseitin graphs of linear treewidth with high probability. Since such Tseitin formulas require exponential resolution refutations, our result follows. We generalize this argument to a new class of formulas that capture a basic form of parity reasoning involving a sum of two random parity constraints with random orders. Even when the variable order for the sum is chosen favorably, these formulas remain hard for resolution. In contrast, we prove that they have short DRAT refutations. We show experimentally that the running time of CDCL SAT solvers on both classes of formulas grows exponentially with their treewidth.

Hardness of Random Reordered Encodings of Parity for Resolution and CDCL | VIDEO

Recently, convolutional neural networks (CNNs) have become the best quantitative encoding models for capturing neural activity and hierarchical structure in the ventral visual pathway. However, the weak interpretability of these black-box models hinders their ability to reveal visual representational encoding mechanisms. Here, we propose a convolutional neural network interpretable framework (CNN-IF) aimed at providing a transparent interpretable encoding model for the ventral visual pathway. First, we adapt the feature-weighted receptive field framework to train two high-performing ventral visual pathway encoding models using large-scale functional Magnetic Resonance Imaging (fMRI) in both goal-driven and data-driven approaches. We find that network layer-wise predictions align with the functional hierarchy of the ventral visual pathway. Then, we correspond feature units to voxel units in the brain and successfully quantify the alignment between voxel responses and visual concepts. Finally, we conduct Network Dissection along the ventral visual pathway including the fusiform face area (FFA), and discover variations related to the visual concept of `person'. Our results demonstrate the CNN-IF provides a new perspective for understanding encoding mechanisms in the human ventral visual pathway, and the combination of ante-hoc interpretable structure and post-hoc interpretable approaches can achieve fine-grained voxel-wise correspondence between model and brain. The source code is available at: https://github.com/BIT-YangLab/CNN-IF.

A Convolutional Neural Network Interpretable Framework for Human Ventral Visual Pathway Representation

Offline reinforcement learning (RL) aims to learn an effective policy from a pre-collected dataset. Most existing works are to develop sophisticated learning algorithms, with less emphasis on improving the data collection process. Moreover, it is even challenging to extend the single-task setting and collect a task-agnostic dataset that allows an agent to perform multiple downstream tasks. In this paper, we propose a Curiosity-driven Unsupervised Data Collection (CUDC) method to expand feature space using adaptive temporal distances for task-agnostic data collection and ultimately improve learning efficiency and capabilities for multi-task offline RL. To achieve this, CUDC estimates the probability of the k-step future states being reachable from the current states, and adapts how many steps into the future that the dynamics model should predict. With this adaptive reachability mechanism in place, the feature representation can be diversified, and the agent can navigate itself to collect higher-quality data with curiosity. Empirically, CUDC surpasses existing unsupervised methods in efficiency and learning performance in various downstream offline RL tasks of the DeepMind control suite.

CUDC: A Curiosity-Driven Unsupervised Data Collection Method with Adaptive Temporal Distances for Offline Reinforcement Learning

We consider the problem of personalizing audio to maximize user experience. Briefly, we aim to find a filter h*, which applied to any music or speech, will maximize the user’s satisfaction. This is a black-box optimization problem since the user’s satisfaction function is unknown. Substantive work has been done on this topic where the key idea is to play audio samples to the user, each shaped by a different filter hi, and query the user for their satisfaction scores f(hi). A family of “surrogate” functions is then designed to fit these scores and the optimization method gradually refines these functions to arrive at the filter ˆh* that maximizes satisfaction. 

In certain applications, we observe that a second type of querying is possible where users can tell us the individual elements h*[j] of the optimal filter h*. Consider an analogy from cooking where the goal is to cook a recipe that maximizes user satisfaction. A user can be asked to score various cooked recipes (e.g., tofu fried rice) or to score individual ingredients (say, salt, sugar, rice, chicken, etc.). Given a budget of B queries, where a query can be of either type, our goal is to find the recipe that will maximize this user’s satisfaction. 

Our proposal builds on Sparse Gaussian Process Regression (GPR) and shows how a hybrid approach can outperform any one type of querying. Our results are validated through simulations and real world experiments, where volunteers gave feedback on music/speech audio and were able to achieve high satisfaction levels. We believe this idea of hybrid querying opens new problems in black-box optimization and solutions can benefit other applications beyond audio personalization.

Sample-Constrained Black Box Optimization for Audio Personalization | VIDEO

Shuffled regression is the problem of learning regression models from shuffled data that consists of a set of input features and a set of target outputs where the correspondence between the input and output is unknown. This study proposes a new deep learning method for shuffled regression called Shuffled Deep Regression (SDR). We derive the sparse and stochastic variant of the Expectation-Maximization algorithm for SDR that iteratively updates discrete latent variables and the parameters of neural networks. The effectiveness of the proposal is confirmed by benchmark data experiments.

Shuffled Deep Regression

Convolutional neural networks (CNNs) and Transformer-based networks have recently enjoyed significant attention for various audio classification and tagging tasks following their wide adoption in the computer vision domain.
Despite the difference in information distribution between audio spectrograms and natural images, there has been limited exploration of effective information retrieval from spectrograms using domain-specific layers tailored for the audio domain. In this paper, we leverage the power of the Multi-Axis Vision Transformer (MaxViT) to create DTF-AT (Decoupled Time-Frequency Audio Transformer) that facilitates interactions across time, frequency, spatial, and channel dimensions.
The proposed DTF-AT architecture is rigorously evaluated across diverse audio and speech classification tasks, consistently establishing new benchmarks for state-of-the-art (SOTA) performance. Notably, on the challenging AudioSet 2M classification task, our approach demonstrates a substantial improvement of 4.4% when the model is trained from scratch and 3.2% when the model is initialised from ImageNet-1K pretrained weights. In addition, we present comprehensive ablation studies to investigate the impact and efficacy of our proposed approach. The codebase and pretrained weights are available on https://github.com/ta012/DTFAT.git

DTF-AT: Decoupled Time-Frequency Audio Transformer for Event Classification

The imperative task of revising or updating the knowledge stored within large language models arises from two distinct sources: intrinsic errors inherent in the model which should be corrected and outdated knowledge due to external shifts in the real world which should be updated. Prevailing efforts in model editing conflate these two distinct categories of edits arising from distinct reasons and directly modify the original knowledge in models into new knowledge. However, we argue that preserving the model's original knowledge remains pertinent. Specifically, if a model's knowledge becomes outdated due to evolving worldly dynamics, it should retain recollection of the historical knowledge while integrating the newfound knowledge. In this work, we introduce the task of Temporal Knowledge Editing (TKE) and establish a benchmark AToKe (Assessment of TempOral Knowledge Editing) to evaluate current model editing methods. We find that while existing model editing methods are effective at making models remember new knowledge, the edited model catastrophically forgets historical knowledge. To address this gap, we propose a simple and general framework termed Multi-Editing with Time Objective (METO) for enhancing existing editing models, which edits both historical and new knowledge concurrently and optimizes the model's prediction for the time of each fact. Our assessments demonstrate that while AToKe is still difficult, METO maintains the effectiveness of learning new knowledge and meanwhile substantially improves the performance of edited models on utilizing historical knowledge.

History Matters: Temporal Knowledge Editing in Large Language Model

Network Pruning is a promising way to address the huge computing resource demands of the deployment and inference of Large Language Models (LLMs). Retraining-free is important for LLMs' pruning methods. However, almost all of the existing retraining-free pruning approaches for LLMs focus on unstructured pruning, which requires specific hardware support for acceleration. In this paper, we propose a novel retraining-free structured pruning framework for LLMs, named FLAP (FLuctuation-based Adaptive
Structured Pruning). It is hardware-friendly by effectively reducing storage and enhancing inference speed. For effective structured pruning of LLMs, we highlight three critical elements that demand the utmost attention: formulating structured importance metrics, adaptively searching the global compressed model, and implementing compensation mechanisms to mitigate performance loss. First, FLAP determines whether the output feature map is easily recoverable when a column of weight is removed, based on the fluctuation pruning metric. Then it standardizes the importance scores to adaptively determine the global compressed model structure. At last, FLAP adds additional bias terms to recover the output feature maps using the baseline values. We thoroughly evaluate our approach on a variety of language benchmarks. Without any retraining, our method significantly outperforms the state-of-the-art methods, including LLM-Pruner and the extension of Wanda in structured pruning. The code is released at https://github.com/CASIA-IVA-Lab/FLAP.

RLfOLD: Reinforcement Learning from Online Demonstrations in Urban Autonomous Driving

Downloads

Next from AAAI 2024

Reward Penalties on Augmented States for Solving Richly Constrained RL Effectively

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES