Since language models (LMs) now outperform average humans on many challenging tasks, it is becoming increasingly difficult to develop challenging, high-quality, and realistic evaluations. We address this by examining LM capabilities to generate code for solving real scientific research problems. Incorporating input from scientists and AI researchers in 16 diverse natural science sub-fields, including mathematics, physics, chemistry, biology, and materials science, we create a scientist-curated coding benchmark, SciCode. The problems naturally factorize into multiple subproblems, each involving knowledge recall, reasoning, and code synthesis. In total, SciCode contains 338 subproblems decomposed from 80 challenging main problems, and it offers optional descriptions specifying useful scientific background information and scientist-annotated gold-standard solutions and test cases for evaluation. OpenAI o1-preview, the best-performing model among those tested, can solve only 7.7\% of the problems in the most realistic setting. We believe that SciCode demonstrates both contemporary LMs' progress towards realizing helpful scientific assistants and sheds light on the building and evaluation of scientific AI in the future.

Accelerometer is the predominant sensor used for low-power context detection on smartphones. Although low-power, accelerometer is orientation and position-dependent, requires a high sampling rate, and subsequently complex processing and training to achieve good accuracy. We present an alternative approach for context detection using only the smartphone's barometer, a relatively new sensor now present in an increasing number of devices. The barometer is independent of phone position and orientation. Using a low sampling rate of 1 Hz, and simple processing based on intuitive logic, we demonstrate that it is possible to use the barometer for detecting the basic user activities of IDLE, WALKING, and VEHICLE at extremely low-power. We evaluate our approach using 47 hours of real-world transportation traces from 3 countries and 13 individuals, as well as more than 900 km of elevation data pulled from Google Maps from 5 cities, comparing power and accuracy to Google's accelerometer-based Activity Recognition algorithm, and to Future Urban Mobility Survey's (FMS) GPS-accelerometer server-based application. Our barometer-based approach uses 32 mW lower power compared to Google, and has comparable accuracy to both Google and FMS. This is the first paper that uses only the barometer for context detection.

A numerical simulation model of shallow sea topography imaging by Along-Track Interferometric SAR (ATI-SAR) is introduced according to its imaging mechanism. This simulation model is used to study the influence of radar parameters (frequency, incidence angle, baseline length and polarization) on one dimensional sand wave imaging by ATI-SAR. The simulation results indicate that high frequency, VV polarization, an incidence angle of 50° and long baseline length may be favorable to shallow sea topography imaging by ATI-SAR when the imaging time lag between two antennas is shorter than the de-correlation time of the sea.

This paper introduces a new technique, called peer-assisted rendering, that aims to enable interactive navigation in a 3D networked virtual environment using a resource-constrained device, by speeding up the rendering. A resource-constrained client requests part of the rendered scenes from other peers with similar viewpoints within the virtual environment, and merges the rendered parts into its own view. This approach is more scalable than previous solutions based on server-based pre-rendering. The goal of this paper is to make a strong case for the feasibility of peer-assisted rendering through the following two messages. First, by analyzing a large number of user traces from a popular virtual world called Second Life, we show that there are surprisingly many users with similar viewpoints and encompass large number of common objects in their viewing areas, indicating that a client can potentially find multiple other peers that can assist in rendering. Second, by combining three different rendering methods, each contributing to rendering of different classes of objects in the scene, we show that it is possible for a client to render the scene efficiently with little visual artifacts.

Synthetic aperture radar (SAR) is a major sensor to study internal waves (IWs). In this paper, we numerically simulate IWs in SAR images to investigate the optimal observation conditions of the IWs. In the simulation process, two SAR parameters are considered simultaneously, and different wind fields are also taken into account. Several SAR images from previous experiments (e.g., Surveillance Satellite Project, Synthetic Aperture Radar Internal Wave Signature Experiment, etc.) are referred to validate some of our simulation results. The optimal observation conditions are explored to provide useful guidance to observation and inversion of IWs in SAR images.

For wireless sensor networks, collaboration of multi-sensors (acoustic, magnetic, radiation, mechanical) enhances the ability of detection. In this paper, an effective approach is proposed in order to process collaboratively acoustic and magnetic data from heterogeneous nodes in underwater sensor networks. And theoretic analyze and simulation of this method are shown in this paper.

To detect the distributed targets in SAR images of the sea, an algorithm based on local spectral histograms (LSH) and agents is proposed. The filter banks consist of the intensity filter and 36 Gaussian derivative filters at 6 orientations and 3 scales. After picking out the background images of the sea, the distribution of the difference value based on LSH is obtained. Given a probability, a threshold is achieved which will be used in judging behavior of the agent. Then an agent system is proposed, and a group of behaviors are introduced, including judging, moving, communicating, breeding, inheriting and dying behaviors. So the detection could be performed through the evolution of the agents. Several examples show that our algorithm is effective for different distributed targets.

In order to develop bistatic SAR ocean investigation, the first issue is the study of bistatic SAR ocean surface imaging mechanism. And velocity bunching mechanism is a SAR inherent modulation. Thus, the intensity expression in the bistatic SAR image plane is derived in this paper. The expression describes quantificationally the displacement of the scatter elements in the image plane and a degradation in radar resolution in azimuthal direction. In addition, a bistatic SAR linear velocity bunching transfer function is obtained in the linear parameter regime. Finally, this paper analyzes qualitatively the effects that various platform observation conditions have on the bistatic SAR linear imaging range of wave spectrum.

Lyzenga's paper ("Unconstrained inversion of waveheight spectra from SAR images" IEEE Trans. Geosci. Remote Sens., vol. 40, no. 2, pp. 261-270, Feb. 2002) after being corrected (ibid., vol. 40, no. 3, p. 729, Mar. 2002) has still a few remaining errors. This paper presents the additional corrections.

Recently, as the rapid growing use of dual polarization spaceborne SAR data for ship detection, it seems necessary in practice for us to find a favorable detection algorithm, however, which is still scarce in literatures. In this paper, based on the application of the whitening filter, we proposed to design a dual polarization CFAR (Constant False Alarms Rate) detection method for ship detection, with the verification of its CFAR property. Preliminary detection results of real Envisat/ASAR Alternating Polarization (AP) mode data were shown in the last section to validate the detection method.

Bistatic SAR is a potential and effective tool for ocean remote sensing, and the simulation of bistatic SAR raw data from ocean surface plays a great role in the understanding of the mechanism of bistastic SAR ocean surface imaging. In this paper, a bistaitc SAR raw data simulation model for ocean surface is presented, which includes bistatic scattering simulation and ocean echo data simulation. In our research, based on IEM scattering model, a stochastic multiscale scattering model is improved by applying the second order nonlinear modulation model, and the speckle noise and coherent time character of ocean surface are simulated in the SAR raw data simulation.

Optimization of antenna array using in SAR system is considered in this study. A robust evolutionary algorithm, non-dominated sorting genetic algorithms (improved NSGA-II), is applied on a spaceborne SAR antenna pattern design. The system consists of two objective functions with two constraints. Pareto front are generated as a result of multi-objective optimization. After being validated by a test problem ZDT4, the algorithms were used to synthesis antenna radiation pattern. The good results with low ASR and high directivity are obtained.

Detection of distributed targets such as internal wave or ship wake on sea surface in Synthetic Aperture Radar (SAR) images is an important application of ocean microwave remote sensing. The chaotic characteristic of sea clutter gives some clues to targets detection on sea surface. Speckle, the inherent noise of SAR image, will affect the predictive accuracy of sea clutter adversely and reduce the detection performance of radar. In order to apply the chaotic characteristic of sea clutter to targets detection in SAR images more effectively, an improved chaos-based detection method is proposed in this paper. First, speckle noise is suppressed by undecimated wavelet transform (UWT), and then targets are detected on the basis of the chaotic characteristic of sea clutter from the denoised SAR images. Experimental results prove that the method proposed in this paper is effective for the distributed targets detection in SAR ocean images.

A fractional Fourier transform based method for radar imaging of moving targets is presented in this paper. Compared with the traditional joint time-frequency approaches, it can achieve a better result of clutter cancellation by using optimal filtering in the fractional Fourier domain due to linear and rotational properties of this transform. By using this method, we can carry out moving targets detection, LFM signal parameters estimation as well as clutter cancellation at the same time under the time-frequency analysis. The effectiveness of this method has been tested through simulation.