Search Results for author:
Found 31 papers, 11 papers with code
ClevrSkills: Compositional Language and Visual Reasoning in Robotics
Robotics tasks are highly compositional by nature.
Multi-Draft Speculative Sampling: Canonical Architectures and Theoretical Limits
In this work we show that the optimal scheme can be decomposed into a two-step solution: in the first step an importance sampling (IS) type scheme is used to select one intermediate token; in the second step (single-draft) speculative sampling is applied to generate the output token.
AirLetters: An Open Video Dataset of Characters Drawn in the Air
We introduce AirLetters, a new video dataset consisting of real-world videos of human-generated, articulated motions.
Your Context Is Not an Array: Unveiling Random Access Limitations in Transformers
Despite their recent successes, Transformer-based large language models show surprising failure modes.
Live Fitness Coaching as a Testbed for Situated Interaction
The benchmark requires vision-language models to recognize complex human actions, identify possible mistakes, and provide appropriate feedback in real-time.
Unleashing the Creative Mind: Language Model As Hierarchical Policy For Improved Exploration on Challenging Problem Solving
Large Language Models (LLMs) have achieved tremendous progress, yet they still often struggle with challenging reasoning problems.
Painter: Teaching Auto-regressive Language Models to Draw Sketches
In this work, we apply LLMs to image generation tasks by directly generating the virtual brush strokes to paint an image.
Look, Remember and Reason: Grounded reasoning in videos with language models
Multi-modal language models (LM) have recently shown promising performance in high-level reasoning tasks on videos.
Deductive Verification of Chain-of-Thought Reasoning
In light of this, we propose to decompose a reasoning verification process into a series of step-by-step subprocesses, each only receiving their necessary context and premises.
Is end-to-end learning enough for fitness activity recognition?
End-to-end learning has taken hold of many computer vision tasks, in particular, related to still images, with task-specific optimization yielding very strong performance.
Metaphors We Learn By
In this essay we relate parameter sharing (``weight sharing'') to analogy making and the school of thought of cognitive metaphor.
On the effectiveness of task granularity for transfer learning
We describe a DNN for video classification and captioning, trained end-to-end, with shared features, to solve tasks at different levels of granularity, exploring the link between granularity in a source task and the quality of learned features for transfer learning.
The "something something" video database for learning and evaluating visual common sense
Neural networks trained on datasets such as ImageNet have led to major advances in visual object classification.
Ranked #115 on
Action Recognition
on Something-Something V2
Theano: A Python framework for fast computation of mathematical expressions
Since its introduction, it has been one of the most used CPU and GPU mathematical compilers - especially in the machine learning community - and has shown steady performance improvements.
Architectural Complexity Measures of Recurrent Neural Networks
In this paper, we systematically analyze the connecting architectures of recurrent neural networks (RNNs).
Ranked #23 on
Language Modelling
on Text8
Generating images with recurrent adversarial networks
Gatys et al. (2015) showed that optimizing pixels to match features in a convolutional network with respect reference image features is a way to render images of high visual quality.
Regularizing RNNs by Stabilizing Activations
We stabilize the activations of Recurrent Neural Networks (RNNs) by penalizing the squared distance between successive hidden states' norms.
Denoising Criterion for Variational Auto-Encoding Framework
Denoising autoencoders (DAE) are trained to reconstruct their clean inputs with noise injected at the input level, while variational autoencoders (VAE) are trained with noise injected in their stochastic hidden layer, with a regularizer that encourages this noise injection.
How far can we go without convolution: Improving fully-connected networks
We propose ways to improve the performance of fully connected networks.
RATM: Recurrent Attentive Tracking Model
The proposed Recurrent Attentive Tracking Model performs well on all three tasks and can generalize to related but previously unseen sequences from a challenging tracking data set.
Neural Networks with Few Multiplications
For most deep learning algorithms training is notoriously time consuming.
Dropout as data augmentation
Dropout is typically interpreted as bagging a large number of models sharing parameters.
Conservativeness of untied auto-encoders
We discuss necessary and sufficient conditions for an auto-encoder to define a conservative vector field, in which case it is associated with an energy function akin to the unnormalized log-probability of the data.
EmoNets: Multimodal deep learning approaches for emotion recognition in video
The task of the emotion recognition in the wild (EmotiW) Challenge is to assign one of seven emotions to short video clips extracted from Hollywood style movies.
On Using Very Large Target Vocabulary for Neural Machine Translation
The models trained by the proposed approach are empirically found to outperform the baseline models with a small vocabulary as well as the LSTM-based neural machine translation models.
Modeling Deep Temporal Dependencies with Recurrent Grammar Cells""
We propose modeling time series by representing the transformations that take a frame at time t to a frame at time t+1.
Zero-bias autoencoders and the benefits of co-adapting features
We show that negative biases are a natural result of using a hidden layer whose responsibility is to both represent the input data and act as a selection mechanism that ensures sparsity of the representation.
Modeling sequential data using higher-order relational features and predictive training
In this work we extend bi-linear models by introducing "higher-order mapping units" that allow us to encode transformations between frames and transformations between transformations.
Unsupervised learning of depth and motion
We present a model for the joint estimation of disparity and motion.
Learning to encode motion using spatio-temporal synchrony
To this end, we show that the detection of spatial transformations can be viewed as the detection of synchrony between the image sequence and a sequence of features undergoing the motion we wish to detect.
Gated Softmax Classification
We describe a log-bilinear" model that computes class probabilities by combining an input vector multiplicatively with a vector of binary latent variables.
